ABUSIVE SUPERVISION - Sage Publications

ABUSIVE SUPERVISION

Abusive supervision refers to sustained displays ofnonphysical forms of hostility perpetrated by supervi-sors against their direct reports. Examples of behaviorthat fall within the abusive supervision contentdomain include public derogation, undermining, andexplosive outbursts. Key features of the construct arethat abusive supervision refers to ongoing manifesta-tions of hostility rather than discrete episodes and thatabusers may or may not intend to cause harm. Hence,for example, yelling at subordinates for the purpose ofeliciting greater task performance could be consideredabusive. It should also be noted that abusive supervi-sion constitutes a subjective assessment, in the sensethat behavior that is perceived to be abusive in onecontext may not be so perceived in another context,and two subordinates could render different interpre-tations of the same supervisor behavior. Similar con-cepts that have been the focus of systematic empiricalresearch include bullying, petty tyranny, and down-ward mobbing.

EPIDEMIOLOGY

According to epidemiological studies, abusive super-vision is much more common than physical violenceor sexual harassment; one in seven employees reportsthat his or her current supervisor is abusive, approxi-mately 50% of employees can expect to have an abu-sive supervisor at some point in their working life, andmost abusers target multiple subordinates simultane-ously. Half of abusive supervisors are women, most

abusers target same-sex victims, and there are sexdifferences in terms of the ways in which men andwomen abuse their subordinates; women bulliesengage in more social manipulation (i.e., rumors andinsulting comments about one’s personal life), andmale bullies engage in more covert aggression, actsthat on the surface appear rational, such as appraisingtargets unfairly and preventing them from expressingthemselves.

OBSTACLES TO SYSTEMATICEMPIRICAL INQUIRY

There are challenges associated with studying abusivesupervision, not the least of which is the fact thatresearchers typically rely on subjective reports as toindividuals’ level of exposure. A problem with thisapproach to measuring abusive supervision is thatsome people may underreport their level of exposurebecause they are reluctant to admit that they have beenvictimized, whereas others exaggerate their super-visors’ hostility. A related obstacle to conducting validempirical research is that linking abusive supervisionand important outcomes requires gathering data fromabused subordinates who are willing to identify them-selves. Failing that, perceived abuse cannot be linkedwith data collected from independent sources (e.g.,observers, supervisors, archival records). A third chal-lenge is that organizations may be hesitant to allowresearchers to administer surveys on the topic. What isclear is that although abusive supervision is a low-base-rate phenomenon that is difficult to study, theresearch to date consistently suggests that its effectscan be severe.

1

A

A-Rogelberg-45020.qxd 7/12/2006 11:43 AM Page 1

CONSEQUENCES OF ABUSIVE SUPERVISION

Compared with nonabused subordinates, abused sub-ordinates have higher quit rates, and among those whostay in the job, abusive supervision is negativelyrelated to subordinates’ job satisfaction, commitmentto the organization, and trust in the supervisor, andpositively related to psychological distress (i.e.,depression, anxiety, and burnout) and conflict betweenwork and family obligations. The cost per serious caseof abuse in the workplace has been estimated atbetween $17,000 and $24,000 in terms of absen-teeism, turnover, legal costs, and reduced productivity,and the total cost to organizations has been estimatedto be more than $23.8 billion in the United Statesalone.

Abusive supervision is not strongly related tobottom-line measures of productivity such as salesvolume, number of units produced, and work quality.The most likely reason for this is that employees can-not easily modify these kinds of performance contri-butions, regardless of how they might feel about theirboss. For example, assembly-line workers cannot sim-ply stop producing when they do not like something atwork, and salespeople on commission cannot stopselling to get back at their boss, at least not withouthurting themselves. But research suggests that abusedsubordinates will retaliate against their supervisor andtheir organization by withholding citizenship behav-iors, contributions such as being helpful and courte-ous, and showing initiative. Abused subordinates canexpress their resentment by modifying their citizen-ship behavior, because these contributions are to alarge extent discretionary, meaning that they fallbeyond the job requirements. These kinds of contribu-tions are very important because they provide organi-zations with flexibility and the capacity to cope withuncertainty. Hence, organizations may be at a com-petitive disadvantage when a substantial percentage ofsubordinates withhold citizenship because their super-visors are abusive.

Although abused subordinates tend to performfewer acts of citizenship than do nonabused subordi-nates, some abused subordinates will nevertheless doso. However, there are differences in the ways abusedsubordinates respond to their coworkers’ citizenshipperformance. Intuitively, we would expect thatemployees will have more favorable attitudes towardtheir job when their coworkers perform more acts ofcitizenship. This notion is rooted in the assumption that

good citizenship makes the workplace a more attrac-tive and comfortable environment. However, it wasfound that this was not the case for work groups inwhich the supervisor was more abusive. In thoseinstances, employees were less satisfied when theircoworkers engaged in greater citizenship behavior.Subsequent inquiry explained why this was so. Ingroups led by abusive supervisors, subordinates per-formed citizenship behaviors not out of a genuinedesire to benefit the organization, but to portray them-selves in a favorable light, to make their coworkerslook less dedicated by comparison, and to direct theirsupervisors’ hostility at others. Consequently, acts ofcitizenship may cause fellow coworkers to experienceunfavorable attitudes when the supervisor is abusive.

MODERATING FACTORS

Abusive supervision does not affect all employees thesame way. In three studies, it was found that the dele-terious effects of abusive supervision on employees’attitudes and psychological health were more pro-nounced when the subordinate has less job mobility(i.e., when the subordinate is trapped in a job becausehe or she has few attractive alternatives to the currentposition), when the abuse is selective rather than dis-tributed (i.e., when subordinates are singled out forabuse as opposed to being targeted along with others),and when the target attributes the abusive behavior tostable characteristics of the supervisor (e.g., mean-ness, incompetence, or indifference) rather than tocharacteristics of the organization (e.g., time pressuresor competitive work climates).

Another study found that subordinates’ personali-ties influenced how they responded to abusive super-vision. This study suggested that abused subordinateswere more likely to engage in dysfunctional forms ofresistance (i.e., nonconformity to downward influenceattempts that involves outright refusal and ignoringthe supervisor’s requests) and that this effect wasmore pronounced among subordinates who were dis-positionally disagreeable (i.e., unconcerned aboutthe quality of their interpersonal relationships withcoworkers) and dispositionally low in conscientious-ness (unconcerned about fulfilling task-related obli-gations). This research provides support for the ideathat subordinates’ personalities influence the extent towhich they engage in retaliation behaviors againstabusive supervisors; employees retaliate against

2———Abusive Supervision


abusive supervisors by actively refusing to do whattheir supervisors want them to do, but only when theyare unconcerned about the relational and task-relatedconsequences associated with noncompliance.

ANTECEDENTS OF ABUSIVE SUPERVISION

Comparatively little research has explored theantecedents of abusive supervision. One study revealedno consistent relationships between hostile supervisorbehavior and supervisor disposition (e.g., theory Xbeliefs, low self-esteem, and low tolerance for ambi-guity), situational factors (e.g., institutionalizednorms, power, and stressors), or their interactions. Amore promising line of inquiry has taken a victim-precipitation perspective, the notion that some indi-viduals may become at risk of being victimized byeliciting or provoking the hostility of potential perpe-trators and that perpetrator and situational factorscontribute more strongly to the occurrence of abusivesupervision when a vulnerable target is available. Thestudy in question found that supervisors who experi-enced procedural injustice (i.e., decision makers usingunfair procedures during the process of renderingallocation decisions) were more abusive when theyhad a high negative-affectivity subordinate, one whowas dispositionally inclined to experience high levelsof distressing emotions and who was likely to be per-ceived as weak, vulnerable, and ripe for exploitation.An implication of this finding is that supervisorsinclined to hostility choose targets strategically, focus-ing their abuse on subordinates who appear to be“good” targets. This work also suggests that perpetra-tors may express their hostility against targets otherthan the source of their frustration (i.e., subordinateswho are not responsible for the injustices supervisorsexperience).

COPING WITH ABUSIVE SUPERVISION

Is there anything abused subordinates can do to copewith their supervisors’ hostility? Abused subordinatesuse two general kinds of coping strategies, which maybe labeled avoidant coping (physical and psychologi-cal withdrawal, maintaining physical distance, notcoming to work, and reliance on drugs and alcohol)and active coping (directly communicating injusticesto the supervisor). Research suggests that abusedsubordinates are more likely to use avoidant coping

than active coping but that the use of active coping isa more effective strategy; in a 6-month longitudinalstudy, it was found that the relationship betweenabusive supervision measured at Time 1 and psycho-logical distress (i.e., burnout, anxiety, and depression)measured at Time 2 was stronger when subordinatesused avoidant coping and weaker when subordinatesused active coping. That is, active coping buffered thestressful effects of abusive supervision and avoidantcoping exacerbated those effects.

CONCLUDING COMMENTS

Given the significant costs that abusive supervisioncan have for organizations and their members, organi-zations would be well-advised to take it seriously.This involves a two-pronged effort focusing on both(a) prevention and (b) management of abuse when itdoes occur. Organizations can prevent the occurrenceof abusive supervision by fostering a culture of civil-ity that is incompatible with abusive behavior. Thiscan be accomplished by implementing 360-degreefeedback programs and training employees and man-agers to develop the skills needed to provide and toopenly receive constructive feedback. Where abuseoccurs, organizations can manage its effects by devel-oping disciplinary procedures for those who violatethe norms for acceptable interpersonal behavior,encouraging victims and witnesses to come forward,and sending the message that those claims will betaken seriously. The evidence suggesting that directcoping produces more favorable outcomes thanavoidant coping means that it may be fruitful toencourage rank-and-file employees to bring abusivebehavior to the attention of higher authorities. In addi-tion, managers should be trained to spot some ofthe markers for abusive behavior, such as withdrawalbehaviors, low morale, and distrust.

—Bennett J. Tepper

See also Leadership and Supervision; Workplace Incivility

FURTHER READING

Aquino, K., Grover, S. L., Bradfield, M., & Allen, D. G.(1999). The effects of negative affectivity, hierarchicalstatus, and self-determination on workplace victimiza-tion. Academy of Management Journal, 42, 260–272.

Ashforth, B. (1997). Petty tyranny in organizations: A pre-liminary examination of antecedents and consequences.

Abusive Supervision———3


Canadian Journal of Administrative Sciences, 14,126–140.

Einarsen, S. (2000). Harassment and bullying at work: Areview of the Scandinavian approach. Aggression andViolent Behavior, 5, 379–401.

Tepper, B. J. (2000). Consequences of abusive supervision.Academy of Management Journal, 43, 178–190.

Tepper, B. J., Duffy, M. K., Hoobler, J., & Ensley, M. D.(2004). Moderators of the relationships between cowork-ers’ organizational citizenship behavior and fellowemployees’ attitudes. Journal of Applied Psychology, 89,455–465.

Tepper, B. J., Duffy, M. K., & Shaw, J. D. (2001).Personality moderators of the relationships betweenabusive supervision and subordinates’ resistance.Journal of Applied Psychology, 86, 974–983.

Zellars, K. L., Tepper, B. J., & Duffy, M. K. (2002).Abusive supervision and subordinates’ organizationalcitizenship behavior. Journal of Applied Psychology, 87,1068–1076.

ACADEMY OF MANAGEMENT

The Academy of Management is the oldest and largestscholarly professional management association in theworld. Founded in 1936 by professors Charles L.Jamison and William N. Mitchell, the academy wasformed to create and exchange knowledge aboutthe discipline of management. Broadly defined, thisdiverse field ranges from the study of organizationalprocesses, structures, and behaviors to the examina-tion of environmental, cultural, industrial, and eco-nomic factors that affect organizations and theiremployees.

Located at Pace University and the Lubin School ofBusiness in New York, the Academy of Managementis a member-driven association led by a 16-memberboard of governors, elected by voting members of theacademy. Committees and task forces are appointedby the president of the academy to carry out servicesand conduct projects for the association.

The Academy of Management has 24 professionaldivisions and interest groups, and members belong to aminimum of 2 of these groups. The divisions with themost (i.e., more than 2,500) active members includeOrganizational Behavior; Business Policy andStrategy; Organization and Management Theory;Human Resources; Organizational Development andChange; and International Management. Liaisons

communicate between the board and the academy’scommittees, task forces, divisions, and interest groups.

MEMBERSHIP

The Academy of Management comprises more than15,700 members from 90 countries. Its membershipincludes researchers from colleges, universities, andresearch organizations, as well as practitioners whowork in businesses, government agencies, and not-for-profit organizations. Although there are no require-ments to become a member, those interested in joiningthe academy must apply to one of four membershipcategories: (a) academic (e.g., researcher, teacher); (b)executive (e.g., management, consultant); (c) student;and (d) emeritus (i.e., member for at least 10 yearsprior to retirement). A majority of the membership isin the academic category.

MISSION

The mission of the Academy of Management is toenhance the profession of management by advancingthe scholarship and enriching the professional devel-opment of its members. The academy is also dedi-cated to shaping management research and education.Several core values guide the achievement of this mis-sion: (a) to conduct and share high-quality research,teaching, and practical applications of management;(b) to promote and encourage ethical conduct in activ-ities pertaining to management research, teaching,and practice; (c) to foster an environment that respectsdiverse points of view; (d) to encourage members toshare their knowledge of management and new devel-opments in the field with members located in differentinstitutions and countries; and (e) to build opportuni-ties for professional collaboration and networking toadvance the field.

FUNCTIONS

The Academy of Management sponsors a number ofactivities in pursuit of its mission. Each year, theacademy holds an annual meeting, which is attendedby more than 6,000 people. This meeting provides asetting in which to share research and expertise in alldisciplines of management via invited papers, compet-itive paper sessions, panel discussions, symposia,workshops, eminent speakers, and special sessions fordoctoral students.

4———Academy of Management


In addition, the Academy of Management pub-lishes a variety of management-related periodicals.Among these publications are four scholarly journals:(a) Academy of Management Learning and Educa-tion, (b) Academy of Management Journal, (c)Academy of Management Review, and (d) Academy ofManagement Executive. The Academy of Manage-ment also publishes a newsletter and a best papersproceedings CD from its annual meeting. Each divi-sion within the academy disseminates newsletters forits members, as well.

In terms of electronic communications, theAcademy of Management hosts a wide variety of list-servs, in which interested parties can post or viewmessages about the full spectrum of management top-ics. Several online discussion boards are also hostedby the academy.

Another program offered by the academy is acareer development and job placement service. Jobinterviews and career growth workshops are con-ducted at the conference, and a searchable databasesystem is offered for organizations to post positionsand view applicant credentials, and for potential appli-cants to search for available positions. Professionaldevelopment opportunities are also available throughforums on special topics considered timely and impor-tant to members of the academy. Further, in recogni-tion of contributions to the field, the Academy ofManagement sponsors a number of professionalawards for notable achievement.

—Jennifer L. Burnfield

See also American Psychological Association, Associationfor Psychological Science; Society for Industrial andOrganizational Psychology

FURTHER READING

Academy of Management Web site. Retrieved February 28,2006, from http://www.aomonline.org/

ACTION THEORY

Action theory represents a general model of work-related cognition and behavior with implications fora wide range of topics in industrial/organizationalpsychology. Inspired by Lewin’s field theory,American cybernetic models, and Russian and Polish

approaches, German work psychologists initiated thedevelopment of action theory in the late 1960s. As thecore concept of the theory, action is conceived of asgoal-directed behavior. Actions are behavioral unitsoriented toward their own distinct goals, whereasoperations (e.g., movement patterns) are subordinateaction components. As anticipatory cognitive struc-tures, goals guide the action process, because theyfunction as relatively invariant set points for the inter-pretation of feedback. Action theory explains both thesequential ordering and the hierarchical structuring ofaction.

THE ACTION SEQUENCE

Action theory differentiates five phases of the actionsequence: (a) goal development and choosing betweencompeting goals; (b) orientation (i.e., collecting rele-vant information about the task and the conditionsin one’s work environment) and prognosis of futureevents; (c) plan development and selection; (d) execu-tion of the plan and monitoring; and (e) the process-ing of feedback, which in turn influences thedevelopment of subsequent goals. These action stepsare not always taken in the same order (e.g., initialplans may be refined during action execution). Theaction sequence allows for an analysis of the interfacebetween the objective work environment and subjec-tive task representations, because employees’ specificredefinitions of tasks presented by the organization(e.g., to operate a machine) determine their individualgoals and plans (e.g., whether and how to take actionwhen the machine breaks down).

FOUR LEVELS OF ACTION REGULATION

From a structural point of view, actions are organizedhierarchically, because higher-order goals are brokendown into subgoals, and higher levels of consciousintellectual regulation are superordinate to lowerlevels of automatic operations. Recent versions ofaction theory distinguish four levels of action regula-tion, ordered from lowest to highest:

1. Sensorimotor level. Stereotyped and automaticmovement sequences are organized without con-scious attention.

2. Level of flexible action patterns. Ready-made actionschemata that do not require conscious representa-tion are tailored to situationally defined parameters.

Action Theory———5


3. Intellectual level. New actions in a complex environ-ment are consciously regulated.

4. Heuristic level. Metacognitive strategies such asgeneral problem-solving approaches are pursuedeither consciously or automatically.

THE OPERATIVE IMAGE SYSTEM

One’s accumulated knowledge of the relationshipsbetween specific conditions, actions, and results isstored in the so-called operative image system. Thissystem reflects the cognitive base for action regulationand entails long-term representations of schemata orstrategies applicable to action regulation at all fourlevels (e.g., movement-oriented schemata to be regu-lated at the sensorimotor level, and strategies tobe implemented at the intellectual level). Originally,cyclical test–operate–test–exit (TOTE) units, whichimply that action is taken until there is congruitybetween the current state and a criterion, were consid-ered the basic units of action regulation. To reconcilethis classic discrepancy reduction approach with thenotion of discrepancy creation, action theorists haveemphasized the role of goals as desired end states andthe impact of active approaches on the environment.

APPLICATIONS OF ACTION THEORY

Action theory has implications for several domains,including stress, training, job analysis, work design,error management, emotion regulation, competencedevelopment, and personality enhancement. Actiontheorists emphasize socialization processes by consid-ering malleable facets of personality, motivation, andcognitive ability as dependent variables that may beaffected by work action. For example, work environ-ments encouraging forward thinking induce actionstyles such as planfulness (i.e., the detailed develop-ment and persistent implementation of long-rangeplans). New ideas on error management and the func-tion of errors in the learning and training process werealso derived from action theory. Research examiningwhy so-called superworkers produce superior resultswithout spending more time at work revealed that theyengage more frequently in planning and have betteroperative image systems, reflected in greater knowl-edge of error frequencies, the signals indicatingerrors, and the duration and efficiency of differentstrategies of dealing with errors.

Conceptualizing stress as a disturbance of actionregulation, action theory offers a theoreticallygrounded stressor taxonomy, composed of three cate-gories: (a) regulation obstacles (i.e., interruptions andregulation difficulties such as poor visibility or lack ofinformation); (b) regulation uncertainties (e.g., roleambiguity); and (c) overtaxing regulations (e.g., timepressure). Multiple job analysis tools have been devel-oped based on action theory. These tools typicallyprovide a structured assessment of regulatory require-ments and difficulties (e.g., the degree to which thework requires a conscious development and coordina-tion of new plans). The function of emotions for actionregulation, particularly in service work, has also beenanalyzed within an action theory context. Emotionsenable people to continue with the action processdespite barriers and difficulties. Examples are the moti-vation derived from pride in anticipation of goal attain-ment and the role of negative affect in facilitating anobjective assessment of environmental barriers.

THE VALUE OF ACTION THEORY

As an integrative metatheory, action theory illuminatesthe implications of specific cognitive and social psy-chological theories for industrial/organizational issues.For example, studies based on the theories of action-state orientation and self-discrimination revealed thatdistractible state-oriented individuals are less likely toefficiently translate intentions into action and morelikely to falsely redefine external demands as theirown goals. Action theory also helps explain the impactof societal transformations on work activities. Longi-tudinal research based on action theory demonstratedthat increases in the levels of complexity and controlexperienced by East German employees after the coun-try’s reunification enhanced their personal initiative(i.e., organizationally functional forms of self-started,proactive, and persistent behavior).

In conclusion, action theory distinguishes itselffrom most micro industrial/organizational modelsbecause of its scope, its versatility, its theoreticalfoundation in cognitive science, its applicability tovarious facets of everyday work behavior, and itssimultaneous consideration of objective environ-ments, internal mental operations, and observablebehavioral outcomes. By bridging the gaps betweenthe environment and cognition (e.g., via task redefini-tions in the action sequence) and between cognitionand action (e.g., via plans as starting points for

6———Action Theory


action), action theory integrates cognitivist and behav-ioral approaches. Action theory has been described asa way of thinking that leads to a sharper understand-ing of how our cognitive apparatus is used and shapedin the workplace and in relation to the world weinhabit.

—Michael Frese and Johannes Rank

See also Goal-Setting Theory; History of Industrial-Organizational Psychology in Europe and the UnitedKingdom; Job Performance Models; PerformanceFeedback

FURTHER READING

Frese, M., & Sabini, J. (1985). Goal-directed behavior: Theconcept of action in psychology. Hillsdale, NJ:Lawrence Erlbaum.

Frese, M., & Zapf, D. (1994).Action as the core of work psy-chology: A German approach. In M. D. Dunnette, J. M.Hough, & H. C. Triandis (Eds.), Handbook of industrialand organizational psychology (Vol. 4, pp. 271–340).Palo Alto, CA: Consulting Psychologists Press.

Gollwitzer, P. M. (1999). Implementation intentions—Strong effects of simple plans. American Psychologist,54, 493–503.

Hacker, W. (2003). Action regulation theory: A practicaltool for the design of modern work. European Journalof Work and Organizational Psychology, 12, 105–130.

Zapf, D. (2002). Emotion work and psychological strain: Areview of the literature and some conceptual considera-tions. Human Resource Management Review, 12,237–268.

ADVERSE IMPACT/DISPARATETREATMENT/DISCRIMINATIONAT WORK

The phrase to discriminate has two interpretations:(a) to display prejudice toward members of a groupthrough unjustified negative actions, and (b) to mean-ingfully differentiate between people on the basis oftheir characteristics. Discrimination of the secondform involves the ability to ascertain the presence anddegree of characteristics that distinguish one personfrom another. For example, a classical music criticshould discriminate among pianists on the basis oftechnique and interpretation. This interpretation holdsthat it is meaningful to differentiate an exceptional

performer from an average performer on the basis ofrelevant factors. In contrast, discrimination of the firstform invokes notions of preference and social injus-tice. Meaningful differentiation is decidedly absentas people are distinguished based on demographic,nonrelevant factors. Because individuals differ on thebasis of many characteristics, organizations must reg-ularly discriminate between individuals when hiring,allocating resources, and rewarding to effectivelymanage a workforce. When organizations differenti-ate individuals based on job-relevant or organization-relevant factors, this discrimination is meaningful andwarranted. When organizations instead differentiateindividuals on the basis of stereotypes and allow thatdifferentiation to influence decision making, the orga-nization has engaged in workplace discrimination.

PERCEPTUAL FORCESBEHIND DISCRIMINATION

The approximately 80,000 complaints filed annuallywith the Equal Employment Opportunity Commission(EEOC) indicate that employment discrimination iscommon. Sex and racial biases influence performanceappraisals and promotion decisions. Black and Hispanicapplicants tend to receive less favorable evaluationsduring interviews. Interpersonal discrimination—avoiding, distancing, and excluding members of agroup—regularly limits minority group members’access to developmental mentors and networks andinterferes with workplace productivity. Further, reportscontinue to document the presence of a “glass ceiling”that blocks women and racial minorities from gainingleadership positions in organizations.

The perceptual processes theorized to produce dis-criminatory behavior explain part of why discrimina-tion at work persists. For individuals, discriminationoriginates with the need to sort people into personallymeaningful categories as a means of efficiently pro-cessing the myriad perceptual stimuli encounteredeach day. This social categorization is often based onwhether a person is perceived by the individual to besimilar or different, and it is facilitated by stereotypicbeliefs that a person categorized within a specificgroup possesses certain traits purely because he or sheis a member of that group. For example, a female can-didate for a management-level position may be passedover for promotion if the stereotype held by evaluatorsleads them to assume that she will become emotionalwhen under stress. As illustrated, the categorization

Adverse Impact/Disparate Treatment/Discrimination at Work———7


process influences how individuals evaluate and feelabout other people. Persons categorized as membersof one’s own group invoke more favorable reactions,whereas persons categorized as members of anothergroup invoke less favorable responses.

However, the tendency to categorize others doesnot necessarily translate into discriminatory actions.Various individual and contextual variables, such asthe social composition of the workplace, the salienceof group membership, and the presence of organiza-tional and local norms, will either facilitate or impedethe emergence of discriminatory behavior.

EMPLOYMENT DISCRIMINATION LAW

The legal environment also plays an important role inmanaging the occurrence of discriminatory behaviors.Federal, state, and municipal statutes, as well as vari-ous constitutional amendments and executive orders,afford individuals rights and protection in the eventthat they are discriminated against in an employmentsetting or are involved in any way in an employmentdiscrimination suit. The three primary pieces of fed-eral legislation responsible for regulating discrimina-tion at work are the Civil Rights Act, the AmericansWith Disabilities Act, and the Age Discrimination inEmployment Act.

The Civil Rights Act

The Civil Rights Act (Title VII, 1964, and asamended in 1978 and 1991) prohibits organizationswith 15 or more employees from discriminating againstindividuals on the basis of race (e.g., Caucasian,African, and Asian), ethnicity (e.g., Hispanic), color ofskin, national origin (e.g., Mexican), sex (and preg-nancy or pregnancy-related medical conditions), andreligion in all aspects of employment (e.g., hiring,compensation, training,performancemanagement,dis-charge). The act defined a protected group as a classof individuals who are similar on one of these bases.However, protected group membership is context-specific. For example, a male applicant for an accoun-tant position may not be considered a minority groupmember in that context, but he would likely be consid-ered a protected group member if he applied for a flightattendant position. Thus, the appropriateness of identi-fying certain individuals as minorities depends on thedemographic makeup of the applicant pool, currentjobholders, and the relevant labor market.

Title VII explicitly allowed for discriminatingbetween individuals on the basis of a job-related,meaningful reason or in response to a bona fide senior-ity system. The act also required organizations toprovide reasonable accommodation for employees toengage in religious practices, unless doing so wouldpose an undue hardship. The act created the EEOC, thefederal agency charged with enforcing the provisionsof employment discrimination legislation.As amended,the act banned the adjustment of assessment scores,whether gathered for purposes of promotion, training,or selection, on the basis of protected class status; less-ened the impact of a mixed-motive defense for organi-zations wherein legitimate reasons in justification of adiscriminatory action are rendered in conjunction withunlawful reasons; and limited the ability of individualsto allege reverse discrimination in the context of judi-cially approved affirmative action plans.

The Americans With Disabilities Act

The Americans With Disabilities Act (Title I, 1990)prohibits organizations with 15 or more employeesfrom discriminating against disabled individuals,individuals who have a record of a disability, and indi-viduals perceived to be disabled in all aspects ofemployment. Disabled individuals are defined asthose persons who suffer from a physical or mentalimpairment that is not correctable and substantiallylimits at least one major life activity, such as thinkingor standing. For example, a legally blind individualwho requires the assistance of a seeing-eye dog wouldbe considered disabled, but a legally blind individualwhose vision can be corrected by wearing glasseswould not. Essentially, the verification of a disabilitymust go beyond a medical diagnosis to consider howthat condition affects an individual’s daily life.

Organizations are required to provide reasonableaccommodation to disabled individuals who are other-wise qualified to perform the essential (i.e., mostintegral or critical) functions of the job. This maymean, for example, adjusting break times so that a dia-betic may stop work to conduct necessary insulintests. An accommodation that would involve excessiveexpense relative to an organization’s resources, dra-matically alter the job or business in question, violatecurrent organization systems or policies (e.g., senioritysystems), or represent a direct threat to the health andsafety of other workers may be viewed as an unduehardship for the organization and hence deemed

8———Adverse Impact/Disparate Treatment/Discrimination at Work


unreasonable. Thus, the reasonableness of a givenaccommodation is individual-, organization-, job-, andcontext-specific. Although the disabled individual isinitially responsible for requesting accommodation,the organization and the individual are encouraged towork together to identify possible accommodations.

The Age Discriminationin Employment Act

The Age Discrimination in Employment Act (1967)prohibits organizations with 20 or more employeesfrom discriminating against individuals age 40 andolder in all aspects of employment. Similar to Title VII,the act explicitly allowed for discriminating betweenindividuals in the presence of a rational, job-relatedreason. In addition, the act identified a series of policyexemptions wherein an age requirement would beviewed as lawful. These include (a) organization poli-cies that identify a mandatory retirement age of 65 forbona fide executives, (b) state or local statutes thatestablish a mandatory retirement age for police officersand firefighters, and (c) various age restrictions in thecommercial airline industry. The act also prohibits age-based discrimination even when the individualsinvolved are both within the age-based protected class;for example, an individual who is 60 can allege dis-crimination in favor of an individual who is 45.

TYPES OF EMPLOYMENT DISCRIMINATION

When these laws are violated, an individual may seeklegal redress by filing a claim of discrimination withthe EEOC. Assuming the claim has merit, the EEOCwill pursue conciliation with the offending organiza-tion to settle the dispute. If attempts at conciliationfail, the suit will proceed to the court system. Theoverwhelming majority of complaints are resolvedbefore reaching federal court. Complaints that doreach a courtroom proceed through a series of phasesin which the burden of proof is shifted back and forthbetween the plaintiff (the individual allegedly dis-criminated against) and the organization. The processthrough which that burden is met depends on the typeof discrimination alleged.

Disparate Treatment

Disparate treatment occurs when an individual suf-fers intentional discrimination on the basis of his or

her membership in a protected group. For example,after September 11, 2001, the EEOC experienced anincrease in the number of complaints filed by ArabAmericans and Muslims who experienced harassmentor discharge allegedly on the basis of their nationalorigin or religion. Proving that the organization hadintent or motive to discriminate is a central aspect of adisparate treatment lawsuit. Because intent can rarelybe known, disparate treatment must often be inferredon the basis of circumstantial evidence. To establish aprima facie case of discrimination, the plaintiff mustprove that he or she was differentially treated becauseof membership in a protected group. This may beaccomplished by demonstrating that members of aspecific protected group consistently received dispro-portionately unfavorable actions.

Under a pattern-and-practice claim, this maybe accomplished by demonstrating that members ofa specific protected group consistently receiveddisproportionately unfavorable actions. Under theMcDonnell Douglas/Burdine framework, thismay be accomplished by showing that the plaintiffwas adversely treated relative to an individual ofa different group who was otherwise similarlysituated in terms of qualifications and job-relatedcircumstances.

In response, the organization must provide a legit-imate, nondiscriminatory reason for its actions. Theorganization may argue that the plaintiff did not havethe necessary job-related qualifications or display theexpected level of performance. In a case of disabil-ity discrimination, the organization may argue thatthe requested accommodation was unreasonable. Orthe organization may defend the action by statingthat the decision was based on a bona fide occupa-tional qualification (BFOQ). Through BFOQs, orga-nizations may overtly exclude individuals on thebasis of sex, age, religion, or national origin whensuch exclusion is required for business survivaland/or public and personal safety. A BFOQ assumesthat most individuals within a given protected groupwill be unable to execute a central job requirementand that failure to do so will risk the health of theorganization and the broader public. For example, airtraffic controllers may not be older than age 49. Inthe event that the organization successfully putsforth an acceptable justification, the plaintiff mayestablish that the proffered reason is merely a pretextand that discrimination was the true reason behindthe action.

Adverse Impact/Disparate Treatment/Discrimination at Work———9


Adverse (Disparate) Impact

Adverse (disparate) impact is discrimination thatoccurs when members of a protected group are system-atically excluded based on an employment policy orpractice that is neutral on its face. Disparate impactlawsuits do not require proof of intent. Instead, to estab-lish a prima facie case, the plaintiff must provide statis-tical evidence that a particular minority group is beingadversely affected by a specific employment practice.This evidence may come in three different forms:

1. Documenting that pass rates for a decision-making hurdle do not fulfill the four-fifths rule. ThisEEOC rule of thumb states that if the pass rate of aminority group for a particular hurdle is less than 80%of the pass rate for the group with the highest passrate, the comparative rates are different enough towarrant a conclusion of adverse impact. For example,in Griggs v. Duke Power Co. (1971), the plaintiff pro-vided evidence that the organization requirement tocomplete a cognitive ability test as part of the selec-tion process had a pass rate for White applicants of58% (the highest pass rate) relative to a pass rate forBlack applicants of 6%. The pass rate of the minoritygroup was 10% of the majority group rate, and thusfar below the 80% criterion.

2. Providing evidence of a restricted policywhereby individuals are excluded on the basis of arequired characteristic that is associated with mem-bership in a protected group. For example, institutinga minimum weight requirement of 130 pounds fora given position will disproportionately excludefemales, in that females as a population weigh pro-portionately less than males.

3. Conducting a workforce utilization analysiscomparing the percentage of minority group mem-bers within an organization and within a job to thepercentage of minority group members in the rele-vant labor force for this organization and this job. Ifprotected group members are being consistentlyscreened out, they should be underrepresented in theorganization relative to their availability in the labormarket.

In response to statistical evidence documentinga disparity, the organization must prove that theemployment practice is consistent with businessnecessity and/or is necessary for the safe operation of

the business. In most cases, this amounts to demon-strating that inferences drawn based on the employ-ment practice are valid in that the practice allowsthe organization to meaningfully differentiate amongindividuals on the basis of job-relevant knowledge,skills, or abilities. Should the organization be success-ful in offering evidence of business necessity, theplaintiff may argue for the substitution of an employ-ment practice that is equally relevant to the job butless discriminatory.

—Jill E. Ellingson

See also Americans With Disabilities Act; Bona FideOccupational Qualifications; Civil Rights Act of 1964,Civil Rights Act of 1991

FURTHER READING

Covington, R. N., & Decker, K. H. (2002). Employment lawin a nutshell (2nd ed.). St. Paul, MN: West-Thomson.

Dipboye, R. L., & Colella, A. (2005). Discrimination atwork: The psychological and organizational bases.Mahwah, NJ: Lawrence Erlbaum.

Landy, F. J. (2005). Employment discrimination litigation:Behavioral, quantitative, and legal perspectives. SanFrancisco: Jossey-Bass.

Player, M. A. (2004). Federal law of employment dis-crimination in a nutshell (5th ed.). St. Paul, MN:West-Thomson.

AFFECTIVE EVENTS THEORY

Affective events theory (AET) is a theory of affect (thebroader term for emotional experiences, includingemotion and mood) in the workplace. In addition tofocusing on affect, it encompasses cognitions, behav-ior, attitudes, and other crucial psychological con-structs to explain job behavior and performance. Thetheory primarily builds on the already establishedcognitive appraisal models and has gathered supportfrom many areas of study in the field of emotions tocreate a more encompassing theory of work behavior.

Affective events theory proposes that there are twopaths to job behaviors, both of which are at leastpartially influenced by affective reactions to events atwork. However, cognitive processes play an essentialrole in the creation of these reactions. The theorybuilds on past theoretical successes while also adding

10———Affective Events Theory


a few new elements (in particular, the notion of timeis essential to the model, as well as a more detailedexplanation of emotion in the workplace) in explain-ing job behavior.

ASSUMPTIONS OF AET

Affective events theory makes several assumptionsabout the workplace and the constructs that describepeople’s reactions to events that happen there. Thefirst is that job satisfaction is different from affect.Nevertheless, AET also assumes that affect con-tributes to job satisfaction and can be used to help pre-dict job performance. Related to that, AET assumesthat affect influences performance, typically in a detri-mental way because emotion is assumed to drawresources from other areas, such as cognitive process-ing, motivation, and attention, among others.

Another major assumption in the AET framework isthat events happen over time, which changes affectcontinuously. Those events influence a person’s imme-diate affective state but also vary over time as newevents arise. Some events are likely to create positivereactions, others negative, and the degree of intensitywill also vary from event to event. Because affect iscontinuously changing within an individual, its influ-ence on behavior is also continuously changing.

THE STRUCTURE OF AET

Affective events theory proposes the following modelfor predicting workplace behavior. Work environmentfeatures (such as office features) precede work events(such as a meeting), and those work events causeaffective reactions. Dispositions influence the causaltransition from work events to affective reactions,as well as the affective reactions themselves. Thoseaffective reactions then influence affect-driven behav-iors, as well as work attitudes. Work attitudes are alsoinfluenced by the work environment. Work attitudesin turn influence judgment-driven behaviors.

From that model, one can see that AET proposestwo different paths to behavior, both of which are pre-ceded by affective reactions. Affect-driven behaviorsstem directly from affective reactions to events in theworkplace. Judgment-driven behaviors, on the otherhand, are arrived at by a longer route, going fromaffective reactions to work attitudes (which are alsoinfluenced by work environment features) and then to

behavior. However, the starting point for AET is theevent. Within AET, an event is defined as a change inthe environmental circumstances that one is currentlyexperiencing. That change then elicits affect, whichthen can influence behavior directly (affect-drivenbehavior) or go through job attitudes to influencebehavior indirectly (judgment-driven behavior).

Affect-driven behavior is an almost instantaneousreaction to an event. In many cases, affect-drivenresponses happen almost immediately after an eventoccurs. An example might be when, after being yelledat by the boss, an employee quits his or her job with-out any thought in the heat of the moment. Judgment-driven behaviors, on the other hand, go through acognitive evaluation via job attitudes. This is a longerprocess and is usually more deliberate. Referring backto the example, if the employee did not quit immedi-ately but went back to his or her desk and thoughtbriefly about the other components of the job, such ashis or her coworkers and the job tasks, and then fac-tored those considerations into his or her decision andreinterpreted the situation, the result would be ajudgment-driven behavior. This process might or mightnot lead the person to quit. The resulting behaviors ofaffect-driven and judgment-driven processes may notbe different, but the decision process is. As the termsthemselves imply, affect-driven behavior is primarilyinfluenced by immediate emotional reactions to anevent and is typically regarded as a single-step process,whereas judgment-driven behavior is influenced byboth emotion and cognition and is regarded as a two-step process that involves a reinterpretation of the orig-inal event and the emotion associated with it.

APPRAISAL OF EVENTSLEADING TO BEHAVIORS

Cognitive appraisal theories argue that people strive tomake meaning of work events. The meaning of theevents then sets the stage for emotional reactions tothe event. There are many theories on how peopleappraise meaning, but the general idea is that everysituation has a meaning underlying the event andthose meanings are arrived at by a person’s interpreta-tions of the situation. Different appraisals of situationslead to different emotions and then behaviors amongindividuals. Individuals emphasize different appraisalprocesses when assigning meaning to an event, andthat is why individuals can have different emotionalreactions to the same situation.

Affective Events Theory———11


The process of appraising is often regarded as atwo-step model. The first step, usually termed primaryappraisal, includes several mechanisms, but the basicidea is how much an event is congruent or incongru-ent with one’s goals, attitudes, values, and so forth. Ifan event is seen as congruent, it is assigned a positivevalue, and if incongruent, the event is viewed nega-tively. The primary appraisal mechanisms are con-cerned with whether a stimulus has to do with aperson’s well-being, which leads to a basic assign-ment of “good” and “bad” labels. In many instances,the primary appraisals assign enough meaning to thephenomenon to elicit an affective response. Examplesof these affective responses can be positive emotions,such as love and relief, but also include negative emo-tions, such as fright and anxiety. A fuller examplewith workplace behavior consequences might beone’s computer freezing up, which might lead one tohit it out of frustration via primary appraisal, as onlya “bad” label has been placed on the event and thereaction is immediate without cognitive factors con-tributing to the behavior.

Secondary appraisals consist of more cognitivelydriven processes, such as future expectations or mem-ory, in addition to the primary appraisal. Many emo-tions occur only when secondary appraisals take placein conjunction with the primary appraisal. An exampleof a negative emotion that requires both stages is anger.A positive emotion that requires the secondary stage ofappraisal is happiness. In both cases (anger and happi-ness), the emotion is targeted at a specific situation andnot a general affective response, as is the case with pri-mary appraisals. In other words, secondary appraisalslead to the assignment of more complex meaning to theevent; no longer is the situation just “good” or “bad.”Once that greater meaning is assigned to an event, adiscrete emotion then emerges that influences one’sbehavior in conjunction with current job attitudes. Soin the example of a computer freezing up, instead ofhitting it immediately in a pure affective reaction, theperson would pause for a brief moment and the eventwould be evaluated in two stages, first if the event isgood or bad via primary appraisal, and then addingother information to deal with the situation via sec-ondary appraisal. Affective events theory proposes thatif job attitudes are positive, one might not hit the com-puter and would instead take the time to call a techni-cian for help. If attitudes are negative, one might stilljust hit the computer.

The secondary appraisal process that leads tojudgment-driven behavior is more deliberative andrequires individuals to take more time (although itcould be only a few seconds) to assign the value ascompared with primary appraisals and affect-drivenbehavior. Primary appraisals that lead to affect-drivenbehaviors are not completely cognition-free, althoughthey are more automatic reactions. However, if thestrength of the initial appraisal and the ensuingemotional reaction is robust enough, the primaryappraisal and the affect-driven response can last forsome time.

For every event, the possible responses of an indi-vidual to a given stimuli may initially seem endless,but once a person appraises the situation, the behaviorchoices become narrowed down based on the person’saffective reactions. To date, there is little researchon what types of behavior result from the differentpaths. However, by definition, affect-driven behaviorsshould be more impulsive and less controlled thanjudgment-driven behaviors, which consider more fac-tors before a behavior is pursued. Therefore, affect-driven behaviors should disrupt job performancebecause of their potentially more abrasive social nature,as well as their ability to draw cognitive resources.Judgment-driven behaviors also should reduce jobperformance, because they reduce time spent on jobtasks as well as draw mental resources away fromthose tasks.

SUMMARY

Affective events theory is a theory of how events inthe workplace (in particular, those events that changea person’s affect) influence behaviors at work overtime. Affect then influences behavior in two possibleways, the first being a direct cause of affect-drivenbehavior, which is an almost automatic emotionalresponse to an event. The second way behavior isinfluenced by affect is through its influences on cog-nitions and attitudes (in addition to the initial affectiveresponse), which in turn cause judgment-driven behav-ior; this is regarded as a more deliberate response to anevent or series of events.

—Brian Francis Redmond

See also Affective Traits; Emotions; Job Satisfaction;Judgment and Decision-Making Process; Mood

12———Affective Events Theory


FURTHER READING

Fisher, C. D. (2002). Antecedents and consequences ofreal-time affective reactions at work. Motivation andEmotion, 26, 3–30.

Grandey, A. A., Tam, A. P., & Brauburger, A. L. (2002).Affective states and traits in the workplace: Diary andsurvey data from young workers. Motivation andEmotion, 26, 31–55.

Lazarus, R. S. (1991). Emotion and adaptation. New York:Oxford University Press.

Lazarus, R. S., & Cohen-Charash, Y. (2001). Discrete emo-tions in organizational life. In R. Payne & C. Cooper(Eds.), Emotions at work (pp. 45–81). New York: Wiley.

Paterson, J. M., & Cary, J. (2002). Organizational justice,change anxiety, and acceptance of downsizing: Pre-liminary tests of an AET-based model. Motivation andEmotion, 26, 83–103.

Weiss, H. M., & Cropanzano, R. (1996). Affective eventstheory: A theoretical discussion of the structure, causesand consequences of affective experiences at work.Research in Organizational Behavior, 18, 1–74.

Weiss, H. M., Suckow, K., & Cropanzano, R. (1999).Effects of justice conditions on discrete emotions.Journal of Applied Psychology, 84, 786–794.

AFFECTIVE TRAITS

THE CONCEPT OF AFFECTIVE TRAITS

Trait affect is defined as a tendency to respond tospecific classes of stimuli in a predetermined, affect-based manner. Therefore, an affective trait is consid-ered a relatively stable characteristic of personality.There are two general bipolar dimensions of affectiveresponding: trait positive affect (TPA) and trait nega-tive affect (TNA). High TPA is characterized bythe tendency to experience positively activated emo-tions in general, such as excitement, high energy, joy,enthusiasm, and exhilaration. Persons with low TPAhave a general tendency to be lethargic, apathetic, andlistless, but they do not necessarily experience nega-tive affect. High TNA is defined as the tendency toexperience feelings of anger, guilt, fear, annoyance,and nervousness. Low TNA is the other pole of theTNA dimension, characterized by being placid, calm,and contented. The two dimensions, TPA and TNA,are conceptualized as orthogonal or at least separabledimensions, and they show zero to moderate negativecorrelations with each other. This implies that it is

possible to be simultaneously high or low in both TPAand TNA, high in TPA and low in TNA, and viceversa. Combinations between the extremes are possi-ble, too. The term affective traits refers to a person’saverage level or typical amount of a given emotion,whereas affective states are more temporal, situation-bound experiences of moods and emotions.

Both TPA and TNA can be interpreted as the diag-onal coordinates in a circumplex model of affect thatis built on the orthogonal dimensions of activation andpleasantness. High TPA in this model is a combina-tion of high activation and high pleasantness, and highTNA is a combination of high activation and highunpleasantness.

Whereas TPA has been shown to be robustlyrelated with extraversion, TNA has been similarlylinked with neuroticism, two personality factors fromthe five-factor model of personality (Big Five),although the fit is not perfect. As an explanation,Timothy A. Judge and Randy J. Larsen have devel-oped a model for integrating affect with personality,referring to these relationships. They present evidencethat certain personality traits dispose people to bemore or less reactive to hedonic stimuli, and theydemonstrate that other personality traits indirectly dis-pose people to modulate their emotional reactions.Extraversion and neuroticism are considered to repre-sent differential sensitivity to typical TPA and TNAstimuli. High-neuroticism individuals are mainlymotivated to avoid punishment (negative stimuli),whereas high-extraversion individuals are mainly moti-vated to gain rewards (positive stimuli).

Affective traits are genuinely individual-level con-cepts. In a group work context, individual affectivetraits may combine into a group-level affective tonethat in turn is related to experiences and behaviors inthe work group.

MEASUREMENT OF AFFECTIVE TRAITS

Several instruments are available for measuring affec-tive traits. The instrument that is most often used isthe Positive and Negative Affect Schedule (PANAS),developed by David Watson and his coworkers. It com-prises two 10-item scales, one for assessing positiveand one for assessing negative affect. The items refer tothe high-activation aspect of negative and positiveaffectivity, respectively. Because the PANAS scaleslack low-activation markers of negative and positive

Affective Traits———13


affect, they sample only a limited part of the affectcircumplex. The PANAS shows good reliability andhigh discriminant validity with low intercorrelationsbetween the positive and negative affectivity scales.

In addition to direct measures of affective traitssuch as the PANAS, researchers use personality mea-sures, particularly neuroticism and extraversion scales,for assessing TNA and TPA, respectively.

AFFECTIVE TRAITS AND JOB SATISFACTION

Affective dispositions influence the extent to whichpeople are satisfied with their jobs. A recent meta-analysis conducted by Carl J. Thoresen and his asso-ciates extracted an estimated mean populationcorrelation of ρ = .33 between TPA and job satisfac-tion and of ρ = −.37 between TNA and job satisfac-tion. Those correlations indicate a rather modest butnevertheless substantial relationship between traitaffectivity and job satisfaction. There is also evidencefrom longitudinal studies for a predictive value ofTPA and TNA for several aspects of job satisfactionup to 2 years later, as well as correlations of disposi-tional affect in younger years with job satisfaction inolder years.

The underlying processes through which traitaffectivity influences job satisfaction are not wellunderstood. Most studies concerned with trait affec-tivity and job satisfaction are correlation studies anddo not allow one to test for causality. Research hasconcentrated on TNA rather than TPA. Because high-TNA individuals are more sensitive to negative stim-uli, they are likely to react more negatively whenexperiencing negative job events, which consequentlylowers job satisfaction. Furthermore, it is possible thathigh-TNA individuals have a higher threshold for pos-itive stimuli and therefore react with a lower magni-tude to positive events. They may experience theeffects of positive mood-inducing events to a lowerextent or for shorter periods of time than do low-TNAindividuals. There is some evidence for the assump-tion that TPA represents reward-signal sensitivity andTNA represents punishment-signal sensitivity. Forexample, TPA is related to pay satisfaction (i.e., asalient reward), but TNA is not. Additionally, TNAindividuals may dwell on their failures and those ofothers, thus causing negative interpersonal interac-tions with their peers and superiors and lower jobsatisfaction.

AFFECTIVE TRAITS AND JOB PERFORMANCE

Potential relationships between affective traits and jobperformance have been discussed in the context of thehappy-productive worker hypothesis and the powerof being positive. By drawing on expectancy theory,some researches have argued that individuals high onTPA should show higher task performance because oftheir positive expectations about the relationshipbetween effort and performance and between perfor-mance and positive outcomes. In addition, it has beensuggested that TPA should lead to higher goals andmore persistence in the face of obstacles. Moreover,researchers have proposed that TPA is associated withextra-role and citizenship behaviors, whereas TNAimpedes supportive social interactions.

Although there are many studies on the rela-tionship between trait affect and job satisfaction, farfewer empirical studies have examined the relation-ship between affective traits and job performance.Studies that used rather broad well-being measuresas indicators for affective traits found positiverelationships between an individual’s tendency to experi-ence and show positive affect at work and supervi-sory rating of job performance, also when usinglongitudinal designs. Managers experiencing higherlevels of well-being and positive affect showedhigher decision-making accuracy, higher interper-sonal performance, and higher managerial perfor-mance. In contrast, most studies that used thePANAS to assess trait affect failed to find significantbivariate relationships between TNA or TPA andtask performance. Trait affect has been shown to beempirically related to extra-role performance at theindividual level (e.g., coworker support and workfacilitation) and to prosocial behavior and coopera-tiveness at the group level.

It has been suggested that individual core self-evaluations play an important role for organizationalbehavior. Core self-evaluations comprise self-esteem,generalized self-efficacy, emotional stability (i.e., lowneuroticism), and locus of control. Although thesecore self-evaluations are not affective traits in a nar-row sense, findings on the relationship betweenemotional stability and job performance are relevanthere, because the emotional stability construct largelyoverlaps with TNA. Meta-analytical evidence sug-gests that emotional stability as an aspect of core self-evaluations shows a weak positive correlation with jobperformance.

14———Affective Traits


In addition, meta-analyses on the relationshipbetween personality factors and job performance shedsome light on the relationship between affective traitsand job performance. Neuroticism shows a negativerelationship with various facets of job performance,with most true-score correlations not exceeding ρ = −.20.Extraversion is positively related to job performance,with most true-score correlations staying in the rangebetween ρ = .10 and ρ = .20.

AFFECTIVE TRAITS AND THESTRESSOR–STRAIN RELATIONSHIP

Affective traits, particularly TNA, are related to percep-tions of job stressors and strains, with individuals highon TNA reporting higher levels of job stressors andstrains.These relationships imply that the observed cor-relation between self-reported job stressors and strainsmay be partially caused by TNA. Therefore, it has beensuggested that researchers should statistically controlfor TNA when analyzing relationships between self-reported job stressors and strain. However, this viewhas been challenged in a lively debate in which it hasbeen argued that TNA plays a substantive role in thestressor–stain relationship.

CONCLUSION

There is broad empirical evidence that affective traitsare related to job satisfaction. However, the processesunderlying this relationship need further researchattention. Although well-being measures were foundto be related to job performance, the empirical rela-tionships between affective traits and related person-ality concepts, on the one hand, and task performance,on the other hand, are weak. Affective traits, however,seem to be more relevant for contextual performance.Therefore, one might assume that group or organiza-tional performance benefits more from TPA than doesindividual job performance.

—Sabine Sonnentag and Jennifer L. Sparr

See also Emotions; Job Satisfaction; Mood; Stress, Copingand Management

FURTHER READING

Brief, A. P., & Weiss, H. M. (2002). Organizational behav-ior: Affect in the workplace. Annual Reviews ofPsychology, 53, 279–307.

Cropanzano, R., James, K., & Konovsky, M. A. (1993).Dispositional affectivity as a predictor of work attitudesand job performance. Journal of OrganizationalBehavior, 14, 595–606.

Cropanzano, R., Weiss, H. M., Hale, J. M. S., & Reb, J.(2003). The structure of affect: Reconsidering therelationship between negative and positive affectivity.Journal of Management, 29, 831–857.

Judge, T. A., & Larson, R. J. (2001). Dispositional affectand job satisfaction: A review and theoretical extension.Organizational Behavior and Human Decision Making,86, 67–98.

Thoresen, C. J., Kaplan, S. A., Barsky, A. P., & deChermont, K. (2003). The affective underpinnings ofjob perceptions and attitudes: A meta-analytic reviewand integration. Psychological Bulletin, 129, 914–945.

Watson, D., Clark, L. A., & Tellegen, A. (1988). Develop-ment and validation of brief measures of positive andnegative affect: The PANAS scales. Journal of Person-ality and Social Psychology, 54, 1063–1070.

AFFIRMATIVE ACTION

Affirmative action has been one of the most con-troversial public policies of the past 40 years. Aconceptual definition of affirmative action is anymeasure, beyond a simple termination of discrimina-tory practice, adopted to correct for past or presentdiscrimination or to prevent discrimination fromrecurring in the future. In practice, organizationalaffirmative action programs (AAPs) can and do encom-pass a multitude of actions. These actions are shapedby federal, state, and local laws and regulations.Although some educational institutions apply affir-mative action to student admissions and many coun-tries have corresponding laws and regulations, thisentry is limited to workplace affirmative action in theUnited States.

LEGAL ISSUES

Affirmative action law in the United States is jointlydetermined by the Constitution, legislative acts,executive orders, and court decisions. It is complex,incomplete, and open to revision. The Office ofFederal Contract Compliance Programs is responsiblefor developing and enforcing most AAPs, althoughthe Equal Opportunity Employment Commission(EEOC) enforces AAPs in the federal sector.

Affirmative Action———15


A distinction exists between so-called set-asideAAPs and organization-specific AAPs. Set-asideAAPs exist when a pubic organization (e.g., a munici-pality or federal agency) is required to set a goal fordirecting a certain percentage of its budget to qualifiedfirms—typically those owned by members of anunderrepresented group.

In contrast, organization-specific AAPs are createdfor one of three reasons. First, some organizations arerequired by a court order or an EEOC consent decreeto establish an AAP to compensate for illegal discrim-ination. These AAPs are relatively rare. Second, manyorganizations establish AAPs to satisfy regulatoryrequirements. Specifically, the Rehabilitation Act of1973 and the Vietnam Era Veterans’ ReadjustmentAssistance Act of 1974 require certain federal con-tractors to take affirmative action to employ individu-als with disabilities and certain veterans, respectively.Most important, Executive Order 11246, signed byPresident Lyndon Johnson in 1965 and subsequentlyamended, requires federal contractors to take affirma-tive action to eliminate discrimination on the basis ofrace, color, religion, sex, or national origin. Along thesame lines, state and local laws and regulations mayrequire organizations to take affirmative action toimprove the employment opportunities of variousgroups. Third, some organizations establish AAPs ona fully voluntary basis.

Precisely which organizations are required toestablish AAPs and which actions are required, per-mitted, or forbidden varies with the legal basis forthe AAP. Furthermore, actions of state and federalgovernments are limited by the U.S. Constitution,whereas actions of firms in the private sector are con-strained by state and federal legislation (e.g., the CivilRights Acts of 1964 and 1991). The following briefand incomplete description focuses on affirmativeaction as required by Executive Order 11246, becausethat is the primary source of AAPs in the UnitedStates and is the basis of much of the controversy.

Organizations with annual federal contracts of atleast $10,000 are required to take affirmative actionto eliminate discrimination on the basis of race, color,religion, sex, or national origin. They must establishnondiscrimination policies and communicate thosepolicies to employees and applicants. Organizationswith at least 50 employees and contracts above $50,000are further required to perform and report the results ofutilization analyses in which they compare the genderand racial distributions of their workforce to the

relevant labor markets. The relevant labor market forany position includes only those individuals who arequalified for that position and who reside in the recruit-ment area. If the utilization analysis reveals that allgroups are reasonably represented, no further actionsare required. If the utilization analysis reveals that anygroup defined by gender, race, or ethnicity is underrep-resented, the firm must establish flexible goals to elim-inate the underutilization and must make a good faitheffort (i.e., take affirmative actions) to meet thosegoals. Utilization analyses are not required for otherprotected dimensions (i.e., disability, veteran status,religion), so it is impossible to determine whetherunderrepresentation exists along these dimensions.

An important question is which actions are permit-ted when underutilization is revealed. Federal regula-tions strongly emphasize nonpreferential actions suchas the elimination of barriers and the use of targetedrecruitment or training. Because these approachesmay fail to eliminate substantial underrepresentation,some organizations may want to take stronger actions.In so doing, the firm must not violate the constraintsestablished by the Constitution and antidiscriminationlaw. It is clearly illegal to use quotas or to give prefer-ences to unqualified members of the underrepresentedgroup (e.g., through race norming of the selectiontest). Furthermore, Supreme Court decisions havedetermined that any AAP that gives a positive weightto racial minority status is subject to “strict scrutiny.”Such an AAP must be remedial, narrowly tailored,and temporary; must not trammel the rights of others;and must further a compelling governmental interest.Note that the final requirement can be satisfied onlywithin the public sector. Although it has been sug-gested that private-sector organizations might use theeconomic value of diversity to justify positive weight-ing of racial minority status, it is not clear that such anargument would be approved by the Supreme Court.Although positive weighting of gender requires onlyintermediate scrutiny rather than strict scrutiny, itwould still be a risky approach.

EMPIRICAL RESEARCH

As mentioned previously, affirmative action is a con-troversial public policy. The debate regarding whetherit should be eliminated, perpetuated, or expanded iscomplex. For example, philosophical arguments havebeen offered regarding the appropriateness of usingrace-conscious approaches to attain a race-blind

16———Affirmative Action


society. These arguments tend to focus on race-basedaffirmative action and occasionally gender-basedplans; they rarely mention policies that target veter-ans or individuals with disabilities. These debates alsofocus on preferential forms of affirmative action ratherthan the more common, and legal, nonpreferentialforms. Empirical research, in contrast, has focusedon the consequences of affirmative action and onpredictors of attitudes toward affirmative action.

CONSEQUENCES OF AFFIRMATIVE ACTION

For Organizations

A logical analysis shows that affirmative actioncould either help or hurt organizational performance,depending on details of the AAP and on the proceduresused by the organization in the absence of affirmativeaction. Positive effects should occur if the AAPincreases the organization’s access to the labor market(e.g., through intensive recruitment) or decreases dis-crimination against women or racial or ethnic minori-ties. Negative effects should occur if the organizationuses a preferential AAP that supplants a nondiscrimi-natory procedure. In addition, organizations must bearthe costs of administering AAPs. Consistent with thelogical uncertainty, empirical research has failed todemonstrate any substantial effect of affirmative actionon organizational performance.

For Target Groups

A different line of research has assessed the eco-nomic impact of affirmative action on the targetedgroups. Affirmative action appears to have improvedemployment outcomes of targeted groups, but theeffects have varied in complex ways depending onfactors such as the targeted group, geographic region,time period, and type of position. For example, affir-mative action had a substantial positive impact onAfrican Americans in the South between 1965 and1975, presumably because that time and place offereda substantial opportunity for improvement. On theother hand, affirmative action had virtually little or noeffect during the 1980s, perhaps because the Reaganadministration decreased support for the regulatoryagencies and substantially revised those agencies’policies and procedures. Little or no research existson the effects of affirmative action on employmentof individuals with disabilities or veterans. Because

organizations are not required to report employmentstatistics of these groups, such effects would be diffi-cult to document.

There is also some evidence that affirmative actionmay lead to stigmatization of individuals who belongto the AAP target group. The logic is consistent withthe discounting principle of attribution theory. Whentargeted individuals are selected in the context of anAAP, others are uncertain about whether their selec-tion was because of their performance or the AAP.In the absence of affirmative action, this uncertaintydisappears and the individual is assumed competent.Research reveals such stigmatization when observersbelieve or are told that the AAP involves preferences.It can be eliminated or greatly reduced by providingcompelling evidence that the AAP is nonpreferentialor that the selected individual is fully qualified or hasperformed well.

A related stream of research deals with self-stigmatization by target group members. Accordingto the logic outlined above, members of AAP targetgroups may doubt their own competence and conse-quently lose confidence and interest in the job.Although this effect has been observed, almost allsupportive evidence has come from laboratoryresearch in which White female college students areexplicitly told that they have been preferentiallyselected on the basis of their gender. There is little evi-dence for this effect among racial or ethnic minorities,and the effect is absent or much smaller when partici-pants are given clear evidence of their competence orare told their selection was based in part on merit.

For White Males

A final question concerns the impact of affirmativeaction on White males. Although there are manyreports of backlash—opposition by White malesbased in part on the belief that they have been hurt byaffirmative action—there is surprisingly little researchon this question. Logically, the effect should be nega-tive if affirmative action reverses traditional biasesthat favor White males or if preferential forms of affir-mative action replace nondiscriminatory procedures.The limited research that exists reveals such a nega-tive effect. Of course, this assumes a “fixed pie”situation; if implementation of an AAP enhancesorganizational performance because of the increaseddiversity, that increased performance may help allorganization members.

Affirmative Action———17


Attitudes

Perhaps the largest body of empirical research onaffirmative action has dealt with public attitudestoward the policy. This work has assessed the effectsof structural predictors, perceiver variables, and psy-chological mediators of the effects.

The structural predictor that has received the mostattention is AAP strength—the weight given by theAAP to demographic characteristics. Specifically,the public strongly supports AAPs that require onlythe elimination of discrimination. Support decreasessomewhat for AAPs that are designed to enhance tar-get group opportunities—for example, by requiringtargeted recruitment. There is a further drop in supportif the AAP requires selection of underrepresentedgroup members when their qualifications are equiva-lent to those of other applicants. Note that such anAAP would rarely if ever pass legal muster. Finally,there is strong opposition to AAPs that require prefer-ential selection of underrepresented group memberseven when their qualifications are inferior to those ofother applicants. Although such an AAP would beillegal, many scholars who study attitudes towardaffirmative action attitudes have described it in thoseterms, and many people believe preferences are com-mon. Indeed, although most research on AAP strengthhas involved manipulation of the AAP, research onpublic beliefs reveals corresponding effects, so thatindividuals who believe affirmative action merelyrequires the elimination of discrimination have morepositive attitudes than those who believe it involvespreferences.

The only other structural predictor that hasreceived enough attention to merit conclusions is theidentity of the target group. It appears that attitudes, atleast of White respondents, are more negative whenthe AAP is described as targeting African Americansor minorities than when it is said to target women orindividuals with disabilities.

The two perceiver variables that have received themost attention are respondent race and gender. In gen-eral, African Americans report the strongest supportfor affirmative action and Whites the strongest oppo-sition, with Hispanics and Asians reporting intermedi-ate levels of support. However, this race effect ismoderated by AAP strength, increasing in size as theAAP gives greater weight to demographic status. Theeffect of gender on attitudes is much smaller, but ingeneral, women report more support than do men.

Attitudes are also associated with several opinionvariables. Most significantly, opposition increaseswith the respondent’s racial prejudice and sexism. Inaddition, those who subscribe to a conservative polit-ical ideology or who identify with the RepublicanParty report greater opposition than do those who arepolitically liberal or who identify with the DemocraticParty. Opposition also increases with the level of therespondent’s social dominance orientation—an indi-vidual difference variable that represents a generalopposition to equality and support for group-baseddominance. Finally, support for affirmative action isassociated with the belief that the target group experi-ences discrimination and thus that affirmative actionis needed.

Research on psychological mediators finds thatsupport for affirmative action is positively associatedwith anticipated positive effects of the AAP onthe respondent’s personal self-interest and on therespondent’s demographic group. But the strongestassociation of all is with perceived fairness of theAAP—people support AAPs they consider fair andoppose those they consider unfair. As this would sug-gest, providing a justification increases support foraffirmative action, but only if the justification refers tothe value of diversity or the need to make up for pastdiscrimination; simply citing underrepresentationtypically decreases support instead of increasing it.

—David Albert Kravitz

See also Adverse Impact/Disparate Treatment/Discrimi-nation at Work; Attitudes and Beliefs; Banding; CivilRights Act of 1964, Civil Rights Act of 1991; Diversityin the Workplace; Race Norming; Recruitment; SexualDiscrimination; Uniform Guidelines on Employee Selec-tion Procedures

FURTHER READING

Crosby, F. J. (2004). Affirmative action is dead; long liveaffirmative action. New Haven, CT: Yale UniversityPress.

Crosby, F. J., & VanDeVeer, C. (2000). Sex, race, and merit:Debating affirmative action in education and employ-ment. Ann Arbor: The University of Michigan Press.

Doverspike, D., Taylor, M. A., & Arthur, W., Jr. (2000).Affirmative action: A psychological perspective.Huntington, NY: Nova Science.

Edley, C., Jr. (1996). Not all Black and White: Affirmativeaction, race, and American values. New York: Hill &Wang.

18———Affirmative Action


Gutman, A. (2000). EEO law and personnel practices(2nd ed.). Thousand Oaks, CA: Sage.

Holzer, H. J., & Neumark, D. (2000). Assessing affirmativeaction. Journal of Economic Literature, 38, 483–568.

Kravitz, D. A., Harrison, D. A., Turner, M. E., Levine, E. L.,Chaves, W., Brannick, M. T., et al. (1997). Affirmativeaction: A review of psychological and behavioralresearch. Bowling Green, OH: Society for Industrialand Organizational Psychology.

Leiter, S., & Leiter, W. M. (2002). Affirmative action inantidiscrimination law and policy: An overview andsynthesis. Albany: State University of New York Press.

Rubio, P. F. (2001). A history of affirmative action,1619–2000. Jackson: University Press of Mississippi.

Spann, G. A. (2000). The law of affirmative action: Twenty-five years of Supreme Court decisions on race and reme-dies. New York: New York University Press.

AGE DISCRIMINATIONIN EMPLOYMENT ACT

The Age Discrimination in Employment Act (ADEA)of 1967 (amended in 1986) is a U.S. federal law thatprohibits discrimination in employment against indi-viduals who are at least 40 years old. It was enactedby the Congress to promote the employment of olderpeople based on their ability and to prevent intentionaland nonintentional forms of age discrimination. Theact applies to private, public, and foreign companieswith more than 20 workers located in the UnitedStates, as well as to unions and employment agencies.

PROHIBITED PRACTICES

The ADEA makes it unlawful to discriminate againsta person because of his or her age “with respect to anyterm, condition, or privilege of employment.” Thatprohibition applies to such things as hiring, firing, jobassignments, promotions, training opportunities, dis-cipline, and employee compensation. The ADEA cov-ers individuals age 40 and above; a worker who is 39years old at the time of the alleged discrimination isnot entitled to ADEA protection. The ADEA also pro-hibits employer discrimination among older workers.For example, an employer cannot hire a 50-year-oldover a 60-year-old simply because of age.

Although the ADEA restricts the use of age byemployers, it allows age to be taken into account insome situations. For example, in recognition of the

fact that benefits costs may be higher for older work-ers, the ADEA allows employers to provide differentbenefits to older and younger workers if the amountspent on benefits received by older and younger work-ers is the same. The ADEA also recognizes severalgeneral defenses that may provide a legal justificationfor policies or practices that adversely affect olderworkers, as discussed in the following section.

ESTABLISHING ANDDEFENDING ADEA CLAIMS

Violations of the ADEA may be established usingeither the disparate treatment or disparate impact theo-ries of discrimination. The legal defenses that are rele-vant and potentially available to an employer depend onwhich theory of discrimination the plaintiff relies on.

Disparate Treatment

The disparate treatment theory of age discrimina-tion, or intentional age discrimination, requires plain-tiffs (job applicants or employees) to prove that theemployer used age as a factor in an employment deci-sion or action. Examples include the refusal to hireolder workers based on stereotypes about their limitedcapabilities and excluding older workers from certaintypes of training.

In cases where it is established that the employerhas a policy or practice that treats individuals differ-ently based on age, the employer must prove that theage standard it used is a bona fide occupational qual-ification (BFOQ) for the job in question to avoid afinding of illegal discrimination. The BFOQ defenseis narrowly construed and difficult to establish. Theemployer must prove that age is directly related to theability to perform an important aspect of the job thatgoes to the essence of the employer’s business. It isnot enough to merely show that younger workers tendto perform better on the job; it must be shown thatsubstantially all persons over the age limit cannot suc-cessfully perform the job, or that it is highly impracti-cal to assess the relevant ability on an individual basis.Age-based BFOQs are most commonly supported injobs directly involving public transportation or safety,for which there is credible evidence that abilitiesessential to the job diminish significantly with age.

In the more typical case, where there is not anexplicit age-based policy or practice and the employerdenies that age played any role in the challenged

Age Discrimination in Employment Act———19


employment action, the plaintiff must make an initialshowing of intentional age discrimination using eitherdirect evidence (e.g., help-wanted advertising indi-cating age preferences, disparaging age-relatedcomments) or circumstantial evidence. To establish aprima facie case of disparate treatment using circum-stantial evidence in a refusal-to-hire case, the plaintiffmust show that (a) she or he is a member of the pro-tected age class; (b) she or he was qualified for theposition in question; (c) she or he was denied theposition; and (d) someone significantly younger, withsimilar or lesser qualifications, received the positionshe or he was denied.

If the plaintiff establishes the foregoing, theemployer must rebut the circumstantial evidence ofintentional discrimination by producing evidence thatit had a legitimate, nondiscriminatory explanation forits action (e.g., poor job performance, good faithbelief that someone else was more qualified). Ifthe employer is able to provide such a reason, then theburden shifts back to the plaintiff to show that thereason offered by the defendant is a pretext fordiscrimination.

Disparate Impact

Disparate impact age discrimination claims involveemployer policies or practices that appear neutral ontheir face but that have a substantially greater negativeimpact on older individuals when put into effect. Forexample, in several cases, the employer’s use of whatappeared to be age-neutral physical fitness require-ments in hiring decisions were found to have a sub-stantially greater impact in screening out olderemployees.

Even if a plaintiff meets his or her burden of iden-tifying a specific employer policy or practice thatadversely affects older workers, the employer maystill prevail if it can show that its policy or practiceinvolves a reasonable factor other than age (RFOA).The RFOA defense, unique to the ADEA, requiresthe employer to demonstrate that there is a good orrational business reason for the employer policy orpractice. The RFOA defense requires a standard ofjustification that is significantly lower than the BFOQdefense (i.e., an RFOA is much easier to establish)and somewhat higher than the legitimate nondiscrim-inatory reason showing that will rebut a circumstantialprima face case of disparate treatment. Evidence thatthe challenged policy or practice is related to job

performance would be sufficient, but it may not benecessary. For example, in a 2005 ruling (Smith v. Cityof Jackson, Mississippi, the Supreme Court held thatthe employer’s perceived need to offer junior policeofficers salaries that were competitive in the job mar-ket was an RFOA that justified an employer policythat adversely affected older officers.

REMEDIES FOR ADEA VIOLATIONS

A range of remedies are potentially available to suc-cessful plaintiffs in ADEA cases, including reinstate-ment to their old job, employment, back pay, frontpay, promotion, and court costs. In addition, if itis shown that the employer knew that its actionsviolated the ADEA or showed reckless disregard forwhether its actions violated the act, then the courthas discretion to award liquidated damages equal todouble the amount the plaintiff is otherwise owed.Noncompensatory damages (e.g., pain and suffering)are not available.

IMPACT OF THE ADEA

Without question, the ADEA has increased U.S.employers’ awareness of and sensitivity to the use ofjob applicant and employee age in employment deci-sions. Some provisions of the ADEA have had a directand manifest impact on employer practices. For exam-ple, the 1986 amendment to the ADEA has eliminatedthe use of once common age-based mandatory retire-ment policies for all but a relatively narrow group ofemployees (high-level executives and employees inselected occupations in which age is a BFOQ). Thecontinued dramatic growth in the number of lawsuitsalleging ADEA claims suggests that older workershave also become more aware and less tolerant of age-based employment discrimination. Research investi-gating the impact of the ADEA suggests that althoughevidence of differential treatment based on age canstill be found in the American workplace, overall, theADEA has had a positive impact on the employmentprospects of older workers. More specifically, empiri-cal evidence indicates that the ADEA helped boostthe employment levels of older workers, particularlythose aged 60 and over.

—Mark V. Roehling and Lisa M. Finkelstein

See also Adverse Impact/Disparate Treatment/Discriminationat Work

20———Age Discrimination in Employment Act


FURTHER READING

Bennett-Alexander, D. D., & Hartman, L. P. (2004).Employment law for business. New York: McGraw-Hill.

Clark, M. (2005). Court: Workers can sue for unintentionalage bias. HRMagazine, 50(5), 29–32.

Lindeman, B., & Grossman, P. (1997). Employment dis-crimination laws (3rd ed.). Washington, DC: TheBureau of National Affairs.

Neumark, D. (2003). Age discrimination legislation inthe United States. Contemporary Economic Policy, 21,297–317.

Robinson, R. K., Franklin, G. M., & Wayland, R. (2002).The regulatory environment of human resource manage-ment. New York: Harcourt.

Smith v. City of Jackson, Mississippi, No. 03-1160, 544U.S. (2005).

AMERICAN PSYCHOLOGICALASSOCIATION, ASSOCIATIONFOR PSYCHOLOGICAL SCIENCE

The two primary professional psychological associ-ations in the United States are the AmericanPsychological Association (APA) and the Associationfor Psychological Science (APS; formerly called theAmerican Psychological Society).

AMERICAN PSYCHOLOGICALASSOCIATION (APA)

Description of APA

The American Psychological Association wasfounded in 1892 as a scientific and professionalmembership organization for the field of psychologyin the United States. Headquartered in Washington,D.C., APA comprises 53 divisions, covering the spec-trum of psychological specialties (e.g., clinical, exper-imental, developmental, educational, personality, andsocial). The association is a corporation governed bya six-member board of directors and a council ofrepresentatives. Many of the organization’s tasks arecarried out by boards and committees.

APA Membership

With more than 150,000 members, APA is thelargest psychological association in the world.Membership categories include (a) member (i.e.,

doctorate in psychology from an accredited institu-tion); (b) student affiliate (programs for graduate,undergraduate, and high school students); (c) teacheraffiliate (community college or high school teachers);(d) international affiliate (psychologists who live out-side the United States and Canada); and (e) fellow(elected status; members who have demonstratedunusual and outstanding contributions or performancein psychology on a national level).

Mission of APA

The mission of APA is to advance psychology as ascience and profession and as a means of promotinghealth, education, and human welfare. Five mecha-nisms are used to accomplish this mission: (a) broadlyand liberally encouraging all of the psychological dis-ciplines (specialty areas); (b) promoting psychologi-cal research and improving the methods and mannerin which these studies are conducted; (c) improvingthe qualifications and value of psychologists by main-taining high standards for professional conduct,ethics, education, and achievement; (d) maintainingthe utmost standards for professional ethics and con-duct of members of APA; and (e) increasing andspreading knowledge of psychology through a varietyof methods (e.g., meetings, publications, networking,and discussions).

Functions of APA

The association sponsors services to advance theeducation and practice of psychology, including con-tinuing education workshops, an annual convention,and an awards program. Publications of APA includea monthly news magazine (Monitor on Psychology);specialized newsletters; division newsletters; hundredsof books; more than 40 journals; videos; practice-related pamphlets; and the largest online researchdatabase for psychological information in the world(PsycINFO, PsycARTICLES). A major function ofAPA is to advise decision makers in Congress on adiverse range of legislative and regulatory issues (e.g.,aging, crime, terrorism, substance abuse).

ASSOCIATION FOR PSYCHOLOGICAL SCIENCE

Formation of APS in Relation to APA

During the 1970s, many members of APA grewdiscontented with the association’s primary focus on

American Psychological Association, Association for Psychological Science———21


issues pertaining to clinical practice (e.g., health care).This focus led many members to believe that theirneeds and interests as scientific and academic psy-chologists were not being met. Proposals submitted tomake APA the primary association for scientific psy-chologists were not approved by the membership.Consequently, in August 1988, the American Psycho-logical Society (APS) was formed to advance psy-chology as a scientifically oriented discipline. InJanuary of 2006, APS officially changed its name tothe Association for Psychological Science to betterreflect is core mission and values.

Description of APS

The Association for Psychological Science is anational nonprofit membership organization governedby an 11-member board of directors, representing dif-ferent fields of psychology. In contrast to APA, APS isnot divided into specialty areas; rather, it considers thescientific field of psychology as a whole.

APS Membership

There are more than 14,000 members of APS,including academics, researchers, clinicians, teachers,and administrators. Membership types include (a)member (i.e., doctoral degree in psychology or arelated field, or sustained contributions to scientificpsychology); (b) retired; and (c) student affiliate (grad-uate, undergraduate). The association also offers othercategories of membership, which have reduced mem-bership rates (e.g., first-year PhD, postdoctorates).

Mission of APS

The mission of APS is to promote, protect, andadvance the interests of scientifically oriented psy-chology in research, application, teaching, and theimprovement of human welfare.

Functions of APS

The association publishes a newsletter and threejournals: Psychological Science, Current Directionsin Psychological Science, and Psychological Sciencein the Public Interest. Each year, APS sponsors a pro-fessional conference. In addition, APS honors partic-ularly notable contributors to the field by awardingthem fellow status and through specific achievement

awards. Further, APS has been actively involved inobtaining legislative and federal support for scientificpsychological research (e.g., increasing visibility forhealth and behavioral research within agencies suchas the National Institutes of Health).

—Jennifer L. Burnfield

See also Academy of Management; Society for Industrialand Organizational Psychology

FURTHER READING

American Psychological Association Web site. RetrievedFebruary 28, 2006, from http://www.apa.org/

Association for Psychological Science Web site. RetrievedFebruary 28, 2006, from http://www.psychologicalscience.org/

Hakel, M. D., & Herring, L. (2000). Association forPsychological Science. In A. E. Kazdin (Ed.), Encyclo-pedia of psychology (Vol. 1, pp. 154–157). Washington,DC, and New York: American Psychological Association/Oxford University Press.

Kraut, A. G. (2001). Association for Psychological Science(APS). In W. E. Craighead & C. B. Nemeroff (Eds.), TheCorsini encyclopedia of psychology and behavioralscience (3rd ed., pp. 94–96). New York: Wiley.

AMERICANS WITH DISABILITIES ACT

In 1990, Congress passed the Americans WithDisabilities Act (ADA) to provide equal protectionunder the law to disabled citizens, who are not identi-fied in the Civil Rights Acts of 1964 or 1991 as a pro-tected group. The ADA covered various aspects ofdaily life for the disabled, which are addressed underthe following titles:

Title I: Employment

Title II: Public Services

Title III: Public Accommodations

Title IV: Telecommunications

Title V: Miscellaneous Provisions

This entry considers only Title I, on employment.Title I of the ADA was intended to strengthen the

existing Rehabilitation Act (RA) of 1973 by makinglanguage more specific and by including private-sector

22———Americans With Disabilities Act


employers under the previous umbrella of the RA. Itprovided standards for enforcement of the law andcharged government with the responsibility forenforcement. The ADA is administered by threedifferent agencies: the Department of Justice, forpublic-sector employees; the Equal EmploymentOpportunity Commission (EEOC), for private-sectoremployees; and the Department of Transportation, fornonfederal sectors affecting commerce. Althoughoriginally covering only employers with 25 or moreemployees, the act was amended in 1994 to apply toall businesses with 15 or more employees.

Statistics from the EEOC demonstrate that therights of the disabled are being increasingly defendedthrough litigation. In 1992, ADA claims representedless than 2% of all claims filed with EEOC. From1993 through 2003, these claims have averagedapproximately 20% of all claims filed with EEOC. Toput this figure in context, the respective figure forclaims alleging discrimination on the basis of race orcolor is 36%; on the basis of sex, 31%; and on thebasis of age, 21%. In the year 2003 alone, ADAclaims filed through EEOC resulted in total benefits todisabled claimants of $45.3 million. An ADA claimtypically consists of an individual’s assertion that heor she is disabled, had the necessary qualifications forthe job in question, and was denied an accommoda-tion that would have made it possible to successfullyperform the job, or at least the essential functions ofthat job. The issues of accommodations and essentialfunction are discussed below.

Although race, sex, and age are relatively clearattributes allowing for a simple determination of whois covered, the disability statute is not so easilyapplied. Since the passage of ADA, the U.S. SupremeCourt has heard 11 cases involving it, many of themdealing exactly with issues related to who is coveredby the act. The determination of who is a member ofthe class of disabled depends on several statutory def-initions. A person may be classified as disabled if (a)he or she has a current mental or physical impairmentthat limits a major life activity, (b) can demonstrate ahistory of such an impairment, or (c) can show that heor she is being treated as if he or she has, or is per-ceived to have, such an impairment. But for an indi-vidual’s claim to be covered by the ADA, it is notsufficient to simply demonstrate that he or she has animpairment. This impairment must be shown to sub-stantially limit a major life activity. Major life activi-ties include obvious categories such as self-care,

walking, talking, seeing, hearing, and breathing. Butoften, the category is not so obvious, and the individ-ual may claim that the limited life activity is actuallythat of working. When this is the case, the claimantmust show not merely that he or she is unable to per-form a single job but that he or she cannot performsuccessfully in a broad range of jobs as a result of thedisability. An example from a court case provides anexample of the requirement. An individual with a fearof heights is not seen as substantially limited simplybecause he or she cannot work on the upper floors ofa building, because a wide variety of other jobs areavailable from other employers that do not requireemployees to work in high locations.

In addition to showing that they are disabled usingany of the definitions above, claimants must alsodemonstrate that they are qualified. It is not sufficientfor claimants simply to show that they possess theknowledge, skills, or abilities needed to do the job.Rather, to meet this requirement, a claimant mustshow that he or she has the knowledge, skills, andabilities to perform essential job functions and cansuccessfully perform those essential functions with orwithout a reasonable accommodation. Both of theseconditions must be met before the court will considerany claims of discrimination on the part of theemployer. If the person is not considered disabled(with the added qualification that the disability sub-stantially limits a major life activity) or does not pos-sess the necessary qualifications to perform essentialfunctions of the job, then the case is dismissed.

In a general sense, essential functions define whythe job exists. For example, the essential functions ofa bus or train driver are to guide a vehicle on a pre-scribed route within a fixed period of time and to pickup and discharge passengers. Essential functions of afirefighter are to suppress fires while protecting livesand property. Essential functions are usually identifiedthrough the completion of a job analysis or the exam-ination of a job description that resulted from a jobanalysis. Since 1990, many job analyses have identi-fied certain functions as essential to comply withADA requirements. Various courts have ruled that tomeet the burden of showing that one can perform theessential functions of a job, a claimant must show thathe or she can perform all of the essential functionswith or without an accommodation, not merely someof them. This logic also affects the very definition ofwhat a job is. Claimants have often argued that theycould perform a job if some of the essential functions

Americans With Disabilities Act———23


were moved to other jobs. Courts have held that theemployer is not required to redefine a job for purposesof accommodating a disabled applicant or employee.On the other hand, the employer cannot exclude adisabled applicant or employee from a job becausethe disabled individual cannot perform a nonessentialfunction. As an example, a recent amputee who hadbeen a former patrol officer could ask to be placed ondesk duty in a police precinct. The department mightargue that even desk officers may be required torespond to outside events in an emergency; however,if it could be shown that no such emergency had everoccurred, then the hypothetical essential function ofthe desk-duty officer to respond to outside emergen-cies would be seen as nonessential.

Drug and alcohol abuse occupy special status inADA. Only rehabilitated drug abusers are protected;individuals currently using illegal substances (e.g.,heroin, cocaine, marijuana) are not. The act permitsdrug testing of employees, including former addicts,provided these tests are reasonable. In this context,reasonable means an articulated policy that is nondis-criminatory. For example, a drug testing policy specify-ing that testing will occur after a workplace accidentor as part of a return-to-work determination afteran injury would be seen as nondiscriminatory. In con-trast to those using illegal drugs, those who might bedefined as current alcoholics may be covered by theADA and may request accommodations. However,alcohol abusers are held to the same standard as anyother employee with regard to not consuming alcoholat the work site and not being under the influenceof alcohol when reporting to work. In these instances,alcoholism is not protected as a disability. Neverthe-less, an employer may not take adverse employmentaction against an alcoholic employee because of theconsumption of alcohol during nonwork hours unlessthe consumption has resulted in behavior (e.g., DUI orassault on coworkers during nonwork interactions)that would have led to dismissal or suspension for anyother employee who engaged in similar activities.Appropriate accommodations for alcoholic employ-ees might include (a) making sure the employeeknows about available counseling, (b) asking theemployee to make a commitment to rehabilitation,understanding that failure to honor this commitmentmight result in termination, (c) establishing a ladderof progressive discipline (e.g., verbal warning →written warning → suspension → termination) forthose who continue to drink while in an outpatient

treatment program, and (d) providing the opportunityfor inpatient treatment if outpatient treatment isunsuccessful.

Those individuals diagnosed with AIDS and otherinfectious diseases are also protected by ADA, to theextent that the condition does not pose a direct threatto the health and safety of other individuals. As exam-ples, an HIV-positive surgical nurse who refuses totransfer to a nonsurgical area is not protected by theADA from involuntary reassignment or termination.In contrast, a hospital clerical worker who is HIV-positive cannot be excluded from that position as aresult of some general and nonspecific fear that the dis-ease might be transmitted to patients or coworkers.These examples help to demonstrate the more generalprinciple of context. Diabetics and epileptics mightfunction fine in certain job contexts (e.g., routine officework) yet be considered threats to the health and safetyof others in other contexts (e.g., jobs involving the useof heavy machinery or working in sensitive positions inair traffic control or nuclear power).

The reasonable accommodation requirement ofADA is unique. It means that the employer may berequired to make modifications of the applicationprocess, the work environment, and/or the way inwhich the job functions are performed. It is assumedthat there will be a dialogue between the employerand the disabled applicant or employee that will iden-tify what might be considered a reasonable accommo-dation for the disabled individual. As an example,individuals who are visually impaired may request anoral assessment or someone to read test questions andpossible answers to them. Such individuals, if hired,may request a modified work environment to offerprotection from moving equipment. Finally, they mayrequest various technical devices (e.g., voice recogni-tion equipment, high-power lighting, or magnifica-tion) to enable successful completion of essential jobfunctions. Such accommodations must be reasonableand entail looking at various characteristics of theemploying organization, including the cost of theaccommodation, the financial resources available forsuch an accommodation, and the effect of such anaccommodation on the overall capability of the orga-nization to conduct business.

The ADA also has practical implications for theapplication and employment process. Individuals maynot be asked to disclose information about a disability(other than a request by that individual for an accom-modation in the application process) until after an offer

24———Americans With Disabilities Act


of employment has been made. This is to preventindividuals from being unfairly discriminated againstas a result of a covered disability during the employ-ment screening process. The most obvious point atwhich to run afoul of this protection is the preemploy-ment physical. Although it is permissible to give apreemployment physical to any individual (includingdisabled applicants), it cannot be administered before aconditional offer of employment has been given to suc-cessful applicants. A conditional offer is one that iscontingent on passing a physical examination. Further,such physicals must be administered to all applicants,not simply to those who appear to be disabled or whohave asked for an accommodation. Even at early stagesof the application process—before a conditional offerof employment is made—an employer is permitted toask individuals if they think that they can performessential functions that are described to them. Askingapplicants to undergo testing for illegal drug and alco-hol use as part of the application process does notviolate the provisions of the ADA.

W. F. Cascio, in 2005, suggested several ways inwhich employers may embrace the spirit of ADA.These include the following:

• Making the workplace more accessible to individualswith various physical impairments by installingramps for individuals in wheelchairs or with visualimpairments and installing TTY (teletypewriter) andvoice amplifiers for individuals with hearing impair-ments. Newly designed keyboards and computershave been developed for quadriplegics and individu-als with cerebral palsy.

• Creating a position within the equal employmentopportunity domain of an organization for an indi-vidual who would focus on disability issues. Such aposition would include responsibility for the orienta-tion and socialization of newly hired disabled work-ers. This orientation would include the supervisorsand coworkers of the disabled employee.

• Educating senior managers in disability issues andgaining commitment to recruit, select, and accom-modate individuals with disabilities when necessary.

• Analyzing jobs with the specific aim of identifyingtasks and functions for which various disabilities arenot an impediment to successful performance.

• Describing successful accommodation experiencesto the employees within the organization as well asto those outside of the organization.

The ADA is a complex statute that is still evolving.It was written in a way that encouraged cooperation

and dialogue between an employer and a disabledapplicant or employee. The courts look favorably onevidence of good faith efforts by both parties to workout a reasonable accommodation where possible. Itis best for neither the applicant/employee nor theemployer to forget the importance of this dialogue inan eventual judicial decision, should such a decisionbecome necessary.

—Frank J. Landy

See also Civil Rights Act of 1964, Civil Rights Act of 1991;Drug and Alcohol Testing

FURTHER READING

Cascio, W. F., & Aguinis, H. (2005). Applied psychology inhuman resource management. Upper Saddle River, NJ:Pearson/Prentice Hall.

Gutman, A. (2000). EEO law and personnel practices(2nd ed.). Thousand Oaks, CA: Sage.

Papinchock, J. M. (2005). Title I of the Americans WithDisabilities Act: The short but active history of ADAenforcement and litigation. In F. Landy (Ed.), Employ-ment discrimination litigation: Behavioral, quantitative,and legal perspectives (pp. 294–335). San Francisco:Jossey-Bass.

APPLICANT/TEST-TAKER REACTIONS

The term applicant reactions is used to refer to anapplicant’s affect, attitudes, and cognitions toward aselection process. Applicant reaction models suggestthat reactions are very complex and involve percep-tions of multiple aspects of specific tests and the test-ing process in general. Stephen Gilliland was one ofthe first researchers to put forth a theoretical model ofapplicant reactions, and this model has guided muchof this research over the past decade. Gilliland’smodel is based on theories of organizational justice.Organizational justice is concerned with the fairnessof the distribution of organizational outcomes (out-come fairness) and the fairness of procedures usedto distribute these outcomes (procedural justice).Gilliland adapted the basic principles of organiza-tional justice to provide a comprehensive model ofhow applicants perceive and react to selection proce-dures. This model has received considerable support.

Glliland’s model suggests that selection systemsand tests are viewed favorably by applicants

Applicant/Test-Taker Reactions———25


(i.e., are considered fair) to the extent they complywith or violate procedural and distributive justicerules. These procedural and distributive justice rulesare standards that applicants hold for how they expectto be treated and how selection procedures should beadministered and used. These justice rules determineperceptions of process and outcome fairness, such thatwhen the rules are satisfied, the selection process andoutcome are perceived as fair, but when they are vio-lated, the selection process and outcome are perceivedas unfair. As will be discussed, applicant perceptionsof the fairness of a selection process can influence anumber of important individual and organizationaloutcomes. It should be noted that according toGilliland’s model, justice rules would not directlyrelate to applicant intentions or behavior, but woulddo so indirectly through process fairness perceptions.For example, perceived job relatedness is an exampleof a procedural justice rule. Perceived job relatednessrefers to the extent to which the applicant perceivesthat the content of a test reflects the content of the job(e.g., the knowledge, skills, and abilities required bythe job). Perceived job relatedness has been recog-nized as the most important procedural justice rulebecause it consistently influences fairness perceptionsand, through fairness perceptions, test performance.

Over the years, several researchers have modifiedand expanded Gilliland’s original applicant reactionsmodel to include a number of additional antecedentsand moderator variables. For example, Ann-MarieRyan and Robert Ployhart revised the Gilliland modeland included an applicant’s affective and cognitivestates during the selection processes, as well as gen-eral perceptions about testing and selection, as impor-tant in understanding antecedents and consequencesof applicant reactions.

JUSTICE RULES

In applicant reaction models, procedural and distribu-tive justice rules are important antecedents of fairnessperceptions. Although a number of procedural anddistributive justice rules exist, Gilliland specified 10procedural and 3 distributive justice rules, and thesehave received research attention:

Procedural Justice Rules

1. Job-relatedness. The extent to which a test appears tomeasure content relevant for the job

2. Opportunity to perform. The extent to which appli-cants perceive that the test or test process allowsthem the opportunity to express themselves prior to aselection decision

3. Reconsideration opportunity. The opportunity tochallenge or modify the decision-making process

4. Consistency of administration. The extent to whichselection procedures are used consistently acrossapplicants

5. Feedback. The extent to which applicants receivetimely and informative feedback

6. Selection information. The extent to which appli-cants are informed how the test and selection proce-dures will be used and why they are used

7. Honesty. The extent to which recruiters and testadministrators are truthful and honest in their com-munication with applicants

8. Interpersonal effectiveness of administrator. Theextent to which applicants are treated with respectand warmth from the test administrator

9. Two-way communication. The extent to which appli-cants have the opportunity to offer input and to havetheir views on the selection process considered

10. Propriety of questions. The extent to which questionson tests are appropriate and not offensive

Distributive Justice Rules

1. Equity. The extent to which applicants perceive thatthe outcome of the selection process (whether theyare hired or not) is based on competence or merit

2. Equality. The extent to which applicants, regardlessof knowledge, skills, and abilities, have an equalchance of being hired for the job

3. Needs. The extent to which job offers are distributedon the basis of individual needs (e.g., preferentialtreatment for a subgroup)

CONSEQUENCES OF APPLICANT REACTIONS

Applicant reactions toward selection procedures havebeen found to affect a number of important outcomes,both directly and indirectly. It has been shown thatwhen applicants react positively toward a test, they aremore likely to accept a job offer from the company,recommend the company to others, reapply for a jobwith the company, and perform well once they areemployed by the company. It has also been suggested

26———Applicant/Test-Taker Reactions


that negative applicant reactions may result in a greaternumber of employment lawsuits and a decreased prob-ability an applicant will buy the company’s products inthe future.

One of the most important consequences of appli-cant reactions is the effect reactions have on applicanttest performance. However, this research has almostexclusively examined the effects of applicant reac-tions on cognitive ability test performance and hasneglected the effects of reactions on other test mea-sures. This research has shown that when applicantsreact favorably to a cognitive ability test, they are morelikely to perform well on the test, although the effectsare modest.

REACTIONS TOWARDDIFFERENT SELECTION MEASURES

Initial applicant reactions research focused on com-paring reactions to different types of measures. Forexample, research suggests that reactions towardassessment centers and work simulations tend to bemore favorable than paper-and-pencil tests (e.g., cog-nitive ability measures). The reasoning is that assess-ment centers and work simulations appear to be morejob-related and therefore result in more favorablereactions on the part of the test taker. Further, researchsuggests that personality measures tend to be per-ceived less favorably than other types of selectionmeasures.

Although tests seem to differ in the reactions theyevoke, research suggests that reactions toward testscan be altered in several ways. For example, researchhas shown that making a test more job-related willresult in more favorable applicant reactions. That is,by ensuring that the content of the test (regardlessof test type) reflects the content of the job, one canincrease the likelihood that applicants will respondfavorably to the test. Further, research suggests thatproviding an explanation for why the test is used canmake reactions toward the test more favorable, as canmaking selection decisions in a timely manner.

TEST-TAKING MOTIVATION

Test-taking motivation is an important componentin all applicant reactions models. One of the mostimportant and researched outcomes of applicant reac-tions is test performance, and research has clearlyshown that test-taking motivation partially mediates

the relationship between applicant reactions and testperformance. It has been found that when applicantshave favorable reactions toward a test or testing pro-cess, they perform better on the tests.

More recently, researchers have sought to deter-mine precisely how motivation mediates the relation-ship between applicant reactions and test performanceby considering the multidimensional nature of moti-vation. Based on an established theory of motivation,VIE (valence–instrumentality–expectancy) theory, amultidimensional measure of test-taking motivationhas been developed. The three components of VIEtheory are defined as follows. Valence is the desirabil-ity or attractiveness of an outcome. Instrumentality isthe belief that a behavior will lead to a specified out-come. Expectancy is the subjective probability thateffort will lead to a specified outcome. In a testingcontext, valence refers to the value one places on get-ting the job for which one is taking the test, instru-mentality is the belief that good test performance willlead to one getting the job, and expectancy is theexpectation that one will do well on the test if one putseffort into doing well. Early results suggest that thesethree dimensions of test-taking motivation are dis-tinct, as they demonstrate different relationships withtest performance and applicant reactions.

PRE- AND POSTTEST REACTIONS

Some research has examined both pre- and posttestreactions and how time of measurement influencesrelationships. Pretest reaction measures are adminis-tered before the applicant takes the test or takes part inthe selection process in question. Posttest reactionmeasures are administered after the applicant hastaken the test or been through the selection process.Research generally finds that responses to pre- andposttest reaction measures are similar but not identi-cal. Therefore, researchers have tried to understandprecisely why pre- and posttest measures are some-times different.

In particular, the self-serving bias may explain howapplicants respond to posttest reactions and motiva-tion. Specifically, if applicants have already taken atest, their perceptions of how they performed mayinfluence their reported test reactions and test-takingmotivation. Those who believe they did poorly on thetest may be inclined to blame the test and report thatthey have negative test reactions or indicate that theydid not even try to do well on the test (i.e., they report

Applicant/Test-Taker Reactions———27


low test-taking motivation). Attributing one’s negativeperformance to lack of effort or to a problematic testmay help protect one’s self-esteem. Given these find-ings, it is important for researchers to be aware thatpre- and posttest reaction measures may result indifferent outcomes.

RACE DIFFERENCES INAPPLICANT REACTIONS

Racial differences in applicant reactions exist, withBlacks and Hispanics being more likely to have nega-tive reactions than White individuals. It is believedthat these race differences in applicant reactions maycontribute to race differences in test performance. Inparticular, it is well documented that White individu-als, on average, score substantially higher on cogni-tive ability tests than Black and Hispanic individuals.It is believed that differences in applicant reactionsmay contribute to the differences between how Whitesand minorities perform on cognitive ability tests.Therefore, considerable research has focused on howapplicant reactions may affect the race–test perfor-mance relationship. Research has shown that race pre-dicts test reactions, test reactions predict test-takingmotivation, and test-taking motivation influences testperformance. Thus, race differences on tests may belarger when minority reactions are negative becauseminorities will have lower test-taking motivation andhence lower test performance. Although researchshows that reactions indirectly account for significantvariance in race–test performance relationships, appli-cant reactions do not account for the majority of racedifferences in test performance.

PRACTICAL IMPLICATIONS OFAPPLICANT REACTIONS RESEARCH

As noted earlier, applicant reactions have a number ofimportant consequences. Therefore, test administra-tors and human resource professionals would be wiseto make applicant reactions to selection proceduresas favorable as possible. This is especially true whenan organization is trying to meet diversity goals.Research suggests that minorities tend to have lessfavorable reactions toward selection procedures thanmajority group members. Therefore, minorities will bemore likely to self-select out of the selection process oreven be less inclined to take a job if one were offered.Research also suggests that the more qualified job

applicants are likely to be most influenced by how theyperceive the selection process. Thus, ensuring thatselection procedures are viewed favorably by appli-cants may have the added benefits of increasing minor-ity representation in the selection process and retainingthe most qualified job applicants.

To increase the chances that tests are perceivedfavorably by applicants, organizations can ensure thetests they use are job-related, provide explanations forwhy the test is being used (e.g., the test administratorcan provide information about the validity of the mea-sure), explain how the selection process will proceed(e.g., clearly explain the stages of the selectionprocess), provide feedback to applicants in a timelymanner, and treat applicants consistently and withrespect throughout the selection process. Doing somay result in more favorable reactions.

—Lynn A. McFarland

See also Individual Differences; Organizational Justice

FURTHER READING

Anderson, N., Born, M., & Cunningham-Snell, N. (2002).Recruitment and selection: Applicant perspectives andoutcomes. In N. Anderson & D. S. Ones, Handbook ofindustrial, work, and organizational psychology: Vol. 1.Personnel Psychology (pp. 200–218). Thousand Oaks,CA: Sage.

Gilliland, S. W. (1993). The perceived fairness of selectionsystems: An organizational justice perspective. Academyof Management Review, 18, 694–734.

Gilliland, S. W., & Chan, D. (2002). Justice in organi-zations: Theory, methods, and applications. InN.Anderson & D. S. Ones, Handbook of industrial, work,and organizational psychology: Vol. 2. OrganizationalPsychology (pp. 143–165). Thousand Oaks, CA: Sage.

Ployhart, R. E., & Harold, C. M. (2004). The applicantattribution-reaction theory (AART): An integrative the-ory of applicant attributional processing. InternationalJournal of Selection and Assessment, 12, 84–98.

Ployhart, R. E., Ziegert, J. C., & McFarland, L. A. (2003).Understanding racial differences in cognitive abilitytests in selection contexts: An integration of stereo-type threat and applicant reactions research. HumanPerformance, 16, 231–259.

Ryan, A. M., & Ployhart, R. E. (2000). Applicants’ percep-tions of selection procedures and decisions: A criticalreview and agenda for the future. Journal of Manage-ment, 26, 565–606.

Truxillo, D. M., Bauer, T. N., Campion, M. A., & Paronto,M. E. (2002). Selection fairness information and

28———Applicant/Test-Taker Reactions


applicant reactions: A longitudinal field study. Journalof Applied Psychology, 87, 1020–1031.

Truxillo, D. M., Bauer, T. N., & Sanchez, R. J. (2001).Multiple dimensions of procedural justice: Longitudinaleffects on selection system fairness and test-takingself-efficacy. International Journal of Selection andAssessment, 9, 336–349.

ARMY ALPHA/ARMY BETA

The United States entered World War I late in the con-flict and faced the problem of turning large numbers ofoften poorly educated draftees into an effective army ina short period of time. The American PsychologicalAssociation volunteered its services to the war effort,and a committee, headed by Robert Yerkes and includ-ing psychologists such as Arthur Otis and LewisTerman, was assigned the task of developing a practicalmethod of measuring the intellectual level of individualsin large groups. Their efforts led to the development oftwo tests, Army Alpha and Army Beta. Army Alpha wasa written test that could be administered to large groupsof recruits and that provided a rough measure of generalintelligence. Army Beta, a nonverbal test designed forilliterates and for recruits who spoke little or no English,could also be administered to groups and used simplepictorial and nonverbal instructions.

Army Alpha was made up of 212 true–false andmultiple-choice items, divided into eight subscales:(a) oral directions, which assessed the ability to followsimple directions; (b) arithmetical problems; (c) prac-tical judgment problems; (d) synonym–antonymitems; (e) disarranged sentences, which required sub-jects to rearrange fragments into complete sequences;(f) number series completion, which required exam-inees to infer and complete patterns in series of num-bers; (g) analogies; and (h) information, a generalknowledge subtest. The most basic purposes of ArmyAlpha were to determine whether recruits could readEnglish and to help in assigning new soldiers to tasksand training that were consistent with their abilities.Several of the scales and test formats developed byYerkes and his colleagues for Army Alpha are fore-runners of tests still in use today.

Many draftees were unable to respond to writtentests, because of their limited literacy or their limitedcommand of English; Army Beta was developed toassess the abilities of these examinees. The instructions

for the Beta test were given in pantomime, usingpictures and other symbolic material to help orientexamines to the tasks that made up this test. Army Betaincluded seven subscales: (a) maze, which requiredlooking at a graphic maze and identifying the path tobe taken; (b) cube analysis, which required countingcubes in the picture; (c) X-O series, which requiredreading symbol series to identify patterns; (d) digitsymbol, which required matching digits and symbols;(e) number checking, which required scanning andmatching graphic symbols in numeric forms; (f) pic-ture completion, which required examinees to identifyfeatures required to complete a partial picture; and (g)geometrical construction, which required examinees tomanipulate forms to complete a geometrical pattern.

ADMINISTRATION AND USEOF ARMY ALPHA AND ARMY BETA

The Army Alpha and Army Beta were administered tomore than 1.5 million examinees. Scoring guidelineswere developed with the aim of making Army Alphaand Army Beta roughly comparable. Scores on bothtests were sorted into eight order categories (A, B, C+,C, C−, D, D−, E). Those with the lowest letter gradewere generally considered unfit for service. Examineesreceiving grades of D or D− were recommended forassignment to simple duties, working under closesupervision. Examinees with scores in the middle ofthe test score distribution were recommended for nor-mal soldier duties, whereas men receiving higherscores were recommended for training as noncommis-sioned officers and for officer training.

Army Alpha and Army Beta were perceived as use-ful at the time they were introduced. These tests pro-vided at least a rough classification of men, which wasof considerable utility in making the large number ofselection decisions necessary at that time. The appar-ent success of the army’s group tests did not go unno-ticed in business circles and educational settings.Soon after the war, the demand arose for similar testsin civilian settings; by the mid- to late 1920s, intelli-gence testing was widespread, particularly in schools.

CONTROVERSY OVERARMY ALPHA AND ARMY BETA

The use of psychological tests to make high-stakesdecisions about large numbers of individuals was con-troversial at the time these tests were developed, and

Army Alpha/Army Beta———29


Army Alpha and Army Beta continue to be sources ofcontroversy. First, many of the psychologists whodeveloped these tests were extreme proponents ofhereditarian points of view and often were enthusias-tic supporters of the eugenics movement. Yerkes andhis colleagues used Army Alpha and Army Beta datato argue against immigration and racial mixing, claim-ing that the addition of intellectually inferior races andgroups to the American melting pot was responsiblefor what they regarded as low levels of intelligencein the American population. Psychologists involvedin the development of Army Alpha and Army Betaplayed a prominent role in supporting legislation afterWorld War I that greatly curtailed immigration.

Second, serious doubts were raised about the valid-ity and the utility of both tests, particularly ArmyBeta. Despite efforts to train test administrators, ArmyBeta could be a particularly intimidating and confus-ing experience, and it is unclear whether this test pro-vided useful information. More generally, evidencethat Army Alpha and Army Beta actually contributedto the success of the army in assimilating and trainingthe vast group who were tested is thin. In part, theproblem lies with the fact that the United Statesentered the war so late that the success or failure ofthis test was simply hard to gauge. Army Alpha andArmy Beta were a tremendous administrativesuccess—they allowed the army to quickly processhuge numbers of recruits. However, this set of recruitsbarely had time to receive training and were musteredout of the army shortly after the conclusion of the war.The hypothesis that the use of these tests led to betterdecisions than would have been made using more tra-ditional (largely subjective) methods of classificationsimply could not be tested during World War I. Thedocumented validity and utility of successors to ArmyAlpha and Army Beta suggest that these tests werelikely to make a real contribution, but definitive dataabout the impact of these tests does not exist.

Finally, controversy over Army Alpha and ArmyBeta reflected broader controversy over the value (ifany) of psychological testing in general and intelli-gence testing in particular. Early proponents of psy-chological testing sometimes made extravagantclaims about the value and the importance of thesetests, and there was a substantial backlash against themore sweeping claims about the importance, validity,and implications of tests like Army Alpha and ArmyBeta.

—Kevin R. Murphy

See also Cognitive Ability Tests; Individual Differences;Selection Strategies

FURTHER READING

Jensen, A. R. (1998). The g factor. Westport, CT: Praeger.Schmidt, F. L., & Hunter, J. E. (1998). The validity and

utility of selection methods in personnel psychology:Practical and theoretical implications of 85 yearsof research findings. Psychological Bulletin, 124,262–274.

Waters, B. K. (1997). Army alpha to CAT-ASVAB: Four-score years of military personnel selection and classifi-cation testing. In R. F. Dillon (Ed.), Handbook ontesting (pp. 187–203). Westport, CT: Greenwood Press.

ASSESSMENT CENTER

The assessment center is a skills-evaluation processthat has been used historically in selection and place-ment decision making, in employee skill develop-ment, and more broadly in career development andorganizational succession planning. The process wasinitially developed in the 1930s by the German mili-tary for the primary purpose of officer selection. Themethodology was adopted, based on familiarity withthe German model, shortly thereafter by the Britishmilitary for similar purposes, and subsequently by theAustralian and Canadian militaries.

The assessment center process was first used in theUnited States by the Office of Strategic Services(OSS) during the middle to late years of America’sinvolvement in World War II (1943 to 1945), to helpin the selection of operatives for spy missions. Thepersonality theorist Henry Murray had earlier devel-oped and applied a multiple assessment methodologyin a research project aimed at better understandingpersonality. Subsequently, Murray participated fullyin the creation and implementation of the OSS assess-ment center, which also borrowed heavily from theearlier German and British efforts. For a variety ofreasons, including Murray’s strong influence, thethree-day OSS assessment process was centered onmeasuring relatively holistic personality variables,rather than job-specific competencies.

The first publicized business application of theassessment center methodology took place at AT&Tin the mid-1950s, in what was called the ManagementProgress Study, conceived and led by Douglas Bray.

30———Assessment Center


The classical AT&T assessment center became theprototype of the many business applications that wereto follow. The AT&T process was built directly on theOSS model, but with heavier emphasis on exercisescomprising situational, job-sample tests to assess job-related competencies than on the assessment of holis-tic personality variables.

The AT&T process was aimed at aiding in theselection and development of managers for the com-pany. Several hundred candidates were assessed ingroups of 12 in a three-and-a-half-day process, spreadover a four-year period. Initially, a set of characteris-tics of successful managers, or what we would nowcall management competencies, was identified basedon literature review and the judgment of AT&T inter-nal subject matter experts. Note that no explicit jobanalysis was done to identify these characteristics ofsuccessful managers. Then a series of activities wascreated that would allow the assessors to rate partici-pants in each identified skill area. Activities includeda series of in-basket activities (still a standard featureof management assessment centers), leaderless groupdiscussions (also a centerpiece of the methodology),and a variety of individual and group problem-solvingsituations. The issues dealt with were selected to berealistic ones for the job of manager in the organi-zation. In addition to such high-fidelity job-relatedactivities, several projective and structured personalityinventories were given, participants were interviewedextensively, and each was required to write an autobi-ographical essay. The assessors were largely a mix ofconsultants and psychologists, with some involvementof AT&T incumbent managers, as well.

Since the foundational AT&T study, the assessmentcenter methodology has been refined, researched, andapplied internationally and domestically, in a widevariety of work settings, prominently includinggovernment work (especially police and fire depart-ments), service industries, and industrial settings.These subsequent applications of the assessmentcenter methodology still draw on many of the corefeatures of the seminal AT&T program.

So widespread has the methodology become thatspecific and detailed structural, ethical, and profes-sional guidelines have been established by an organi-zation called the International Task Force onAssessment Center Guidelines. In the most currentversion of those guidelines (the first were codified in1975), 10 essential elements of the assessment centermethodology, which must be present for a process to

be considered an assessment center, are identified, asfollows:

1. Job analysis. Required to establish the critical,relevant, observable performance elements andcompetency categories to be assessed. Competency-modeling procedures may be substituted for classicaljob analysis. The point is to establish through rigor-ous methods the observable behaviors and compe-tencies to be assessed.

2. Behavioral classification. Behaviors demonstratedby participants must be classified and categorizedinto dimensions, skills, competencies, abilities, andso forth.

3. Assessment techniques. Must be designed to revealperformance relative to the key dimensions andcompetencies that are critical in the performance ofthe job.

4. Multiple assessments. A mix of techniques must beselected to allow behaviors revealing of the criticalcompetencies to be observed and assessed.

5. Simulations. The techniques employed must includejob-related simulations, although the assessmentcenter overall need not be limited to job-related simu-lations. Although relatively low-fidelity simulationsmay be adequate for some purposes, especially early-career selection programs for nonmanagement jobs,high-fidelity simulations are preferred, especiallyfor developmental (not only or mainly selection)programs for experienced, high-level incumbents.Acceptable simulations require that the assessee actu-ally demonstrate behaviors, not merely select froma list of multiple-choice options or state intendedactions.

6. Assessors. Multiple assessors observe and assesseach participant, with a typical ratio of 1:2. Assessorsshould not be immediate supervisors of a participant.Diversity (functional, ethnic, organizational level,gender) in the pool of assessors is considered highlydesirable.

7. Assessor training. Assessors must demonstrate com-petence in their role, based on targeted assessor train-ing prior to using those skills. Typically, assessorsshould receive two days of training for each day ofthe assessment process itself.

8. Recording behavior. Assessors may not rely solelyon memory but must document through notes,behavioral/competency checklists, video recording,or some similar method that allows later review.

Assessment Center———31


9. Reports. It is expected that assessors draft prelimi-nary reports, based on their records of the behaviorof assessees.

10. Data integration. Assessors are required to pool theirobservations through some accepted methodology,such that final reports represent the integration of allrelevant data.

The assessment center process sometimes runs foronly a day, or even less, but more commonly is a mul-tiday affair, with the period of two to five consecutivedays being common. Assessees are always put throughthe process in groups (the typical size being 6 to 12 par-ticipants in a center, assessed by three to six assessors,although there are exceptions to these dimensions).Assessors, usually higher-level managers in the com-pany, sometimes aided by psychologists or other con-sultants, typically do not know the participants they areassessing. However, there is some recent evidence tosuggest that center ratings may have higher criterion-related validity when in fact the assessors are familiarwith the participants they are assessing.

Some applications of the methodology aim mainlyor exclusively at selection. When that is the case, it iscommon for participants to receive feedback thatincludes only or mainly the final hiring recommenda-tion and to be given the rationale for that recommen-dation only on request. When the primary goal isdevelopment, however, and whenever the assesseesare current members of the organization, participantsare typically given much more detailed feedback,usually including the final feedback report. Indeed,where assessment centers are part of a broader career-planning and/or succession-planning process, it iscommon for the feedback shared with the participantto be highly detailed and for extensive developmentalplanning and coaching to be included as a key elementin the overall assessment process.

Although the focus of the assessment center ismainly on profiling the competencies of the individ-ual, there are numerous collateral benefits. A well-designed center gives candidates a kind of realistic jobpreview. Also, when an organization has assessed anumber of internal candidates, the organization canget a valuable perspective on where general skillstrengths and deficits are by averaging across the indi-vidual profiles. Thus, for example, if most internalcandidates score poorly on the competency of delega-tion, and the company in question has legitimatelydetermined that such skills are indeed important in the

role being assessed, such data can be used to supporttraining and development programs and other relevantskill-building processes in the organization. Such afinding could also affect the company’s recruiting andhiring strategies. Similarly, longer-range successionplanning is supported by having assessment data on apool of incumbent employees who have the core skillsnecessary for higher-level jobs. In general terms,assessment centers can provide a level of informationthat is valuable to the organization, beyond the skillsand developmental needs of individual candidates.

Since the introduction of the assessment centerconcept into business in the 1950s, the methodologyhas been employed by thousands of organizations.Recent estimates indicate that assessment centers inone form or another are used regularly by as many as2,000 companies in the United States, and perhaps asmany as 70% of large organizations in the UnitedKingdom. They are widespread in industry in otherparts of Europe, as well. Additionally, they are partic-ularly likely to be used by public-sector organizationssuch as local police and fire departments.

There is a huge base of research on the assessmentcenter methodology. Despite some flexibility andvariability in how the methodology is applied (e.g.,number and kinds of activities, duration of the center,level of training of assessors, and even the extent towhich the guidelines are adhered to), the data gener-ally support the conclusion that assessment centerscores show very good criterion-related validity inpredicting job performance in the role being assessed.On average, meta-analyses across many studies showcriterion-related validity estimates for a broad rangeof indicators of performance in the +.35 to +.45 range,and often higher for certain submeasures or for com-posite measures. Also, assessment center results haveshown reasonable validity generalization across dif-ferent jobs within broad job families. It is also oftennoted that compared with other selection and develop-ment approaches, the assessment center methodologyhas very high face or content validity, as judged bycandidates, assessors, and other subject matter expertsalike. Thus, the level of participant acceptance of theprocess is typically high. Further, it is widely claimedthat compared with other selection processes, assess-ment centers are fair. In general, they are found notto discriminate against women, minorities, or othergroups protected by the Equal Employment Opportu-nity Commission. In short, they are so popular becausethey work and are legally defensible. (Note, though,

32———Assessment Center


that there is some recent evidence that older candi-dates may be rated generally lower than youngercandidates, suggesting the possibility of age bias.)

Primary criticisms of the assessment center method-ology center on complexity and cost. Developing andrunning an effective high-fidelity assessment centeris time-consuming and expensive, to the extent thatunless it is going to be run many times, it may be costprohibitive for an organization to develop one. Even ifit is to be run multiple times, each session is time-consuming and expensive in itself, requiring trainingtime for assessors and a significant time commitmentfor assessors, participants, role players, administrators,and so on during the session. Thus, the methodology isused more by large organizations that have large num-bers of managers and potential managers, and wherethe costs of a poor selection or career developmentdecision are great. In such cases, again, the assessmentcenter methodology is widely supported as a valuableselection and development tool.

The lack of construct validity has also been identi-fied as a criticism of the assessment center methodol-ogy. It has been found in terms of convergent andespecially discriminant validity; there is sometimeslittle consistency between a participant’s scores on agiven competency from one exercise to another, andwithin a given exercise, scores on different dimensionsare often highly correlated. Recent research aims atdetermining why assessment centers are so effective interms of their ability to predict job success, despite thelack of demonstrated construct validity.

One current trend of note is the growing use of tech-nology as an aid to the assessment process. Thus, thereis increasing use of video recording of participants’performance during the session, allowing later reviewby assessors and enhancing the feedback opportunitiesfor participants, including self-feedback. Also, there isincreasing use of the computer as an aid in integratingscores across different assessors. Additionally, as com-munications technology has become more widespreadin the workplace, there is less reliance on paper-and-pencil activities and increased use of tools such asvoice mail and e-mail as the more ecologically valid in-basket of today’s workplace.

There is an increase in the integration of variousforms of 360-degree feedback processes (usually abbre-viated compared with conventional stand-alone 360s)into the assessment center feedback package, such thatthose familiar with the work of the assessee contributeto a richer overall feedback experience for participants.

Another growing trend is toward more integratedtotal or continuous simulations, rather than a bits-and-pieces, stop-start approach that takes each exercise asa discrete event. Participants in continuous simula-tions are assigned a role, with relevant backgroundinformation about that role, including a set of roleplayers with whom they may interact in the course ofthe center. Participants and role players then stay inrole for the duration of the session.

To combat the high cost and time investment of theclassical assessment center, there is also a developingtrend toward what is being called ongoing assessment,for career developmental purposes. In this application,internal candidates may meet briefly with assessors atmutually convenient times to do more or less standardassessment center exercises. Rather than conductingthe assessment intensively over a period of a few days,it may be spread over several weeks in and aroundother work responsibilities, thus being accomplishedin a less labor-intensive and costly format.

The methodology is now being increasinglyextended beyond the traditional target group of man-agers and potential managers to include candidatesfor positions such as members of self-directed workteams. Indeed, the need to qualify people for suchexpanded roles (including leadership roles in the flex-ible, high-performance workplace) has become a pri-mary impetus for the application of the methodology.

From its earliest applications in the organizationalsetting, the assessment center has grown to be amajor tool for selection, promotion, and developmentof critical organizational talent around the world.Researchers continue vigorous programs to determinewhy assessment centers work to the extent they do andhow they can be made more effective.

—John Kello

See also Assessment Center Methods; LeadershipDevelopment

FURTHER READING

Bray, D. W., Campbell, R. J., & Grant, D. L. (1974).Formative years in business: A long-term AT&T study ofmanagerial lives. New York: Wiley.

Gaugler, B. B., Rosenthal, D. B., Thornton, G. C., III, &Bentson, C. (1987). Meta-analysis of assessment centervalidity. Journal of Applied Psychology, 72(3), 493–511.

Howard, A. (1997). A reassessment of assessment centers:Challenges for the 21st century. Journal of SocialBehavior and Personality, 12(5), 13–52.

Assessment Center———33


International Task Force on Assessment Center Guide-lines. (2000). Guidelines and ethical considerations forassessment center operations: International task forceon assessment center guidelines. Public Personnel Man-agement, 29(3), 315–331.

Schmitt, N., Gooding, R. Z., Noe, R. A., & Kirsch, M.(1984). Meta-analysis of validity studies publishedbetween 1964 and 1982 and the investigation of studycharacteristics. Journal of Applied Psychology, 69,207–213.

Spychalski, A. C., Quinones, M. A., Gaugler, B. B., &Pohley, K. (1997). A survey of assessment center prac-tices in organizations in the United States. PersonnelPsychology, 50(1), 71–90.

Thornton, G. C., III. (1992). Assessment centers in humanresource management. Reading, MA: Addison-Wesley.

ASSESSMENT CENTER METHODS

The assessment center is a methodology used toselect, promote, and develop people, usually man-agers, in an organization. From the earliest efforts inthe 1930s and 1940s in the military, aimed at selectingofficers and key operatives for highly sensitive mis-sions, to the first systematic application in the organi-zational setting in the 1950s, the assessment centerhas become a familiar tool for skills evaluation.

A defining feature of the assessment center method-ology is that it comprises a battery of tests. Further, thetests must represent a multiplicity of types. No singleinstrument or single type of instrument is sufficient toqualify a skills assessment process as an assessmentcenter, by that name, according to the InternationalTask Force on Assessment Center Guidelines.

The earliest iterations of the assessment centermethodology focused to a significant extent on indi-vidual-differences assessment. The American militaryapplication, for the selection of spy operatives inWorld War II, was influenced strongly by the work ofHarvard psychologist and noted personality theoristHenry Murray and included a good deal of personal-ity profiling. It is still common today for personalityassessment to be incorporated into an overall batteryof assessment center instruments. With the develop-ment of the five-factor model of personality, the NEOtests have become an increasingly familiar part of theoverall assessment center process. There are recentdata suggesting that scores on the conscientiousnessand extraversion scales, in particular, may generally

correlate with overall assessment center scores. At anyrate, especially where the focus is more on develop-ment and less on near-term selection, personalityfeedback is commonly part of the assessment process.

Tests of general cognitive ability are also some-times included as a collateral part of the assessmentcenter process. Much like personality tests, cognitiveabilities tests represent a source of additional feed-back to the assessee, as well as potential predictorsof future work success. Cognitive ability scores docorrelate significantly with overall assessment centerscores, which in turn are broadly predictive of jobsuccess.

Interviews are another familiar support tool of theassessment center methodology. The interviews maybe very general, unstructured ones, aimed at identify-ing the assessee’s background, interests, career goals,and so forth. Or they may be more structured, even sit-uational, in which case they may form integral parts ofthe assessment process. A typical situational interviewwould describe a scenario of the sort an incumbentmight experience at work—say, problems with anupset customer, a conflict between sales and opera-tions, or an employee not following safety procedures.After specifying the situation in detail, interviewerswould ask the assessee how he or she would handlethe situation. The assessee’s answers would then berated in terms of the underlying competencies beingassessed (e.g., relationship management, communica-tion, problem solving).

Beyond personality or intelligence assessment andinterviews, the heart of the assessment center method-ology is a series of individual and group activities inwhich candidates, working individually or in variousgroupings, handle work-related problems and issues.Many of these individual and group exercises werepioneered in the earliest applications of the methodol-ogy and continue to be centerpieces of the assessmentcenter process. Included in this mix of classical meth-ods are the venerable in-basket, leaderless group dis-cussions, and role plays.

The in-basket is sometimes considered the definingtool of the assessment center methodology. The basicidea of the in-basket is that participants are assigned arole, namely the role for which they are being assessed,and are given the kinds of memos, reports, notes, andother communications that an incumbent in the rolebeing assessed might receive in her or his in-basket ona given day. The participants must then act on the in-basket information in the way they would in the

34———Assessment Center Methods


real world, prioritizing issues, writing reply memos,gathering information, calling meetings, and so on. Itis common for the in-basket exercise to be timed.When possible, participants might subsequently beinterviewed after the exercise, to better understand therationale for their responses to in-basket items.

The in-basket provides opportunities to observehow the participant prioritizes and plans, solves prob-lems and makes decisions, and communicates andcoordinates with key resources, for example, all ofwhich are common competencies of the managerialrole. In-basket activities can be considered a kind ofmultidimensional work sample test, to the extent thatthe role being assessed requires just such handling ofincoming communications.

One benefit of the in-basket method is that ittypically includes some documented output by theparticipant. The written recommendation, or marketanalysis, or meeting agenda, or reply memo is a pieceof data that assessors can review and reflect on offlineat their leisure, assisting them in the skills evaluationprocess.

Today, managers spend quite a bit less time than inthe past handling memos in paper form and a muchgreater amount of time in e-mail, voice mail, and cellphone communication. It is increasingly common forthe methodology of high-fidelity assessment centers toreflect current technology. Thus, participants’ work-space in a modern assessment center may include adesktop computer, through which they can receive andsend e-mails, and a phone/voice mail system, throughwhich they can communicate with others. The avail-ability of such technology greatly increases the fidelityof the work setting in the assessment center but alsogreatly increases the workload on the assessors whohave to keep track of such rapid communication viamultimedia for several participants.

The leaderless group discussion is also a familiarclassical feature of the assessment center methodol-ogy. Typically, participants are given an issue to dis-cuss and make recommendations about. As the titlesuggests, no member of the group is formally desig-nated as leader. They may be told, for example, thatthey are all regional sales managers for XYZ Corp.,which is facing a declining market, new competition,and quality problems in production. They are thenasked to meet and fully discuss the issues and makea recommendation to the vice president/generalmanager on how to address the problems. They willhave been given sufficient information (e.g., through

in-basket communications) for them to think strategi-cally about the problems. In some cases, individualparticipants are given different information, or dif-ferent perspectives on the issue at hand, such thatdisagreement and conflict are more likely in thediscussion.

In the course of the leaderless group discussion,participants gravitate to roles in which they are com-fortable. Usually, someone will structure the meetingapproach, someone will keep notes, some bring orga-nized proposals, some brainstorm well, some showskill at developing others’ ideas, and some participatelittle. The format provides an excellent opportunityfor assessors to see a whole range of competenciesrelated to communication, influence, collaboration,resolving disagreements, problem solving, relation-ship management, and the like. To the extent that thecontemporary environment of work puts a premiumon joint, collaborative work and the ability to cooper-ate and compromise as well as provide direction, thistool gives assessors a snapshot of candidates’ interac-tive skills and general approach to teamwork. Theleaderless group discussion has been called the pro-jective test of leadership. If the scenario is realisticand engaging, participants can get caught up in theissue and reveal much about their real approach to col-laborative work. One caution is that strong individualpersonalities can shift the dynamics of a leaderlessgroup such that it is hard to get a valid view of eachindividual’s approach. For example, one extremelycompetitive participant can cause others to becomemore aggressive, or for that matter more compliant,than they normally would be. Assessors must be alertto such artifacts.

In the role-play method, participants are given arole and a scenario to act on with a role player/assessor,whose role is also specified. At the appointed time, theassessee and the role player interact as assessorsobserve and make notes. For example, a participantmay be in the role of a supervisor in a team-based pro-duction environment whose crew has been complain-ing about how the previous shift leaves a mess thatthey have to clean up before they can start runningproduction. They have asked her to talk with the peersupervisor of the off-going shift (played by the roleplayer). The role player is provided with specificguidelines as to how to act and the kinds of themes toweave into the role play (e.g., be a bit defensive—express thoughts such as “I don’t tell you how to runyour shift!”). He is further directed to act realistically

Assessment Center Methods———35


to that role and to base further responses on theapproach the assessee takes to the problem. Finally, heis asked to keep in mind that the point of the role play,as with all of the exercises of the center, is to providea realistic forum in which to see the participant’s skilllevel on key competencies, so he should give the par-ticipant ample opportunity to show active listeningskills, project empathy, be results oriented, and so on.

Beyond the classic “big three,” additional exercisesthat are often found in assessment centers includestructured meetings, presentations, and written analy-ses. In a structured meeting, the participant’s roleis specified. Thus the participant might be a salesmanager leading a meeting with his three sales repsaround a given market issue. An example of a presen-tation might be a 10-minute overview of plant statusby a plant manager, given to his boss (role player), ora sales presentation given by a sales manager to anational client (role player). Written analyses mightinclude completing a budget or doing a cost justificationfor the purchase of a major piece of equipment. Again,as always, such exercises are anchored in the compe-tencies that are being assessed.

Finally, it is normal for some self-assessment (andat times 360-degree feedback) to be included in theoverall assessment center process. Early iterations ofthe methodology had participants write autobiographi-cal statements. More recent applications may directlycall for self-assessment, such as “On a five-point scale,rate your ability in the area of delegation . . . what areyour specific strengths in this area . . . what are yourspecific developmental needs in this area?” Or activi-ties may be built into other exercises that precipitatesuch self-assessment indirectly. Thus, an in-basketmemo from the boss might ask the participant to listwhat he or she sees as the critical competencies of thejob in question and to self-assess and identify personaltraining and development needs in these skill areas.

A major methodological question centers on thelevel of fidelity of the simulations in the assessmentcenter. Especially if the goal of the process is to selectnonmanagerial employees from a pool of outsideapplicants, it is common for the center to be brief(1 day or less) and relatively low-fidelity. Thus activ-ities might be more generic (e.g., a leaderless groupdiscussion about how to motivate employees) thandetailed and specific to the role (a leaderless groupdiscussion about changes to a specific incentive bonusplan, based on information given in the in-basket).With internal candidates, and especially for managerial

positions, it is common for the process to be longer(on the order of 2 to 5 days), more detailed, andhigher-fidelity. Also, under the latter circumstances itis common for the various exercises of the assessmentprocess to be integrated rather than discrete. That is,candidates will be given highly detailed informationabout their role at the outset of the process. Thus, ifthey are in the role of a plant manager in a manufac-turing environment, they will be given specificationssuch as a personal work history (how long they havebeen in that role, where they worked before), theparameters of their plant, organizational charts,detailed information on their division and company,recent history of their plant, background on theiremployees, and key challenges facing their plant.Then from the time the center begins until it is con-cluded, they stay in that role. All the issues that comeup are consistent with the information they have beengiven. They role play with key individuals who havebeen identified in the initial information they havereceived (e.g., their boss, their sales counterpart, keydirect-report employees, the union president). Such anintegrated total simulation approach, if well designed,is typically seen by participants as having especiallyhigh face validity.

Modern technology is not only used to increase thefidelity of the work simulation for participants. It isalso used as a powerful aid to the assessor in the rolesof observer, role player, behavior analyzer, and reportwriter. For example, it is increasingly common forassessors to videotape group discussions, role-playinteractions, presentations, and so on. Then the tapescan be reviewed later for more detailed and thoroughanalysis. Additionally, the videos can be incorporateddirectly into the feedback process, potentially enrich-ing the experience for the participant. Technology canalso be used for the delivery of some elements of theassessment process, including personality profiles orcognitive-abilities assessments, online.

In general, given the complexity and cost of assess-ment centers, there is constant pressure to simplifyand shorten assessment centers without sacrificingfidelity and predictiveness. Thus, there is some move-ment toward the use of online situational interviews inwhich participants choose a multiple-choice responseas their output. Such an item can be taken by an indi-vidual quickly and automatically scored and inte-grated with other scores. In a similar vein, there is amove toward a less formalized overall approach toassessment centers, in which the assessment is done

36———Assessment Center Methods


with internal candidates, typically individually, inbrief interactions over a longer period of time, stillusing assessment center tools such as in-basket activ-ities and role plays. One distinct advantage of suchongoing assessment is that it can be done when thecompany has fewer than the usual 6 to 12 candidatesbut needs to assess competencies in a couple of keyindividuals now. Such truncated methods can be builtto have good criterion-related validity, even thoughthey are not fully consistent with the guidelines of theInternational Task Force on Assessment Centers. Inthe contemporary organizational climate of change, itis expected that such variations on the core methods ofthe assessment center will be increasingly common.

—John Kello

See also Assessment Center

FURTHER READING

Bray, D. W., Campbell, R. J., & Grant, D. L. (1974).Formative years in business: A long-term AT&T study ofmanagerial lives. New York: Wiley.

Byham, W. C. (1970). Assessment centers for spottingfuture managers. Harvard Business Review, 48(4),150–167.

International Task Force on Assessment Center Guidelines.(2000). Guidelines and ethical considerations for assess-ment center operations: International task force onassessment center guidelines. Public Personnel Manage-ment, 29(3), 315–331.

Moses, J. J., & Byham, W. C. (Eds.). (1977). Applying theassessment center method. New York: Pergamon.

Schippmann, J. S., Prien, E. P., & Katz, J. A. (1990).Reliability and validity of in-basket performancemeasures. Personnel Psychology, 43(4), 837–859.

Thornton, G. C., III. (1992). Assessment centers in humanresource management. Reading, MA: Addison-Wesley.

Thornton, G. C., III, & Byham, W. C. (1982). Assessmentcenters and managerial performance. New York:Academic Press.

ATTITUDES AND BELIEFS

Attitudes are the positive or negative evaluations madeabout people, issues, or objects. For example, in anorganizational setting, employees might hold attitudestoward their employer or coworkers, toward work-place issues or regulations, and toward the job itself.

Attitudes form a central foundation of the way thatindividuals think about and come to understand theworld around them; consequently, they influence andare influenced by people’s beliefs and cognitions.Much research has focused on the structure andmeasurement of attitudes, as well as their relation toaffect, beliefs, and behavior. A central question thathas been raised with regard to attitudes is whetherthey are accurate predictors of behavior. Understand-ing processes of attitude formation and change hasalso been a dominant avenue of research.

ATTITUDE STRUCTURE

Attitudes are based on cognitive, affective, and behav-ioral information. Beliefs provide the cognitive basisof an attitude. A belief is the cognitive informationthat one has about an attitude object. For example, aworkplace attitude might be based on beliefs, or cog-nitions, about one’s job. The affective basis of an atti-tude refers to the emotional response that one hastoward the attitude object—for example, the affectthat one feels toward one’s job. The behavioral basisof an attitude refers to actions that are taken withregard to the attitude object, such as job-relatedbehaviors that reflect one’s attitude toward work. Anattitude might be based on any combination of thesethree components. For certain attitudes, componentscan be evaluatively inconsistent with one another. Forexample, a person with an emotionally grueling jobmight experience negative affect toward his or herwork but at the same time hold positive cognitions bybelieving that the job is important and useful. Thisleads to attitudinal ambivalence, which is described asa state of holding both positive and negative evalua-tions of the same attitude object.

The issue of attitudinal ambivalence has receivedrecent attention, reflected in the debate over whetherattitude structure is bipolar or bivariate. Evaluativeprocesses have been traditionally conceptualized asbipolar. According to a bipolar model of attitudes,people’s attitudes can range from very negative (andnot at all positive) to very positive (and not at all neg-ative). This conceptualization implies that negativityand positivity are reciprocal, opposing forces; conse-quently, the more positive one’s attitude is, the lessnegative it will be, and vice versa. One limitation ofthis conceptualization is that it precludes the possibil-ity of attitude ambivalence. To address this issue, analternative conceptualization of attitude structure has

Attitudes and Beliefs———37


emerged in which attitudes are viewed as bivariaterather than bipolar. The bivariate perspective suggeststhat positivity and negativity are separable attitudinalsubstrates, rather than opposite ends of the same con-tinuum; further, each can be separately activated andexert an independent influence on behavior.

ATTITUDE FORMATION

Attitudes form through a variety of processes. Manyattitudes are developed through direct experience withan attitude object or learned through processes ofoperant and classical conditioning. A growing body ofevidence suggests that attitudes may also have agenetic basis.

Direct experience. Attitudes may form through directexperience with a person, issue, or object. Direct inter-action with the attitude object contributes to the forma-tion of a positive or negative evaluation.Attitudes formedthrough direct experience are strong predictors of futurebehavior.

Classical conditioning. When a positive or negative stim-ulus is paired repeatedly with an initially neutral attitudeobject, attitude formation through classical conditioningmay occur. When this occurs, the evaluation paired withthe neutral stimulus eventually becomes associated withthe attitude object itself. Attitude formation through thisprocess often occurs at an unconscious level.

Operant conditioning. Attitudes are formed throughoperant conditioning when an attitude object becomesassociated with a positive or negative consequence.Specifically, when behavior toward an attitude object isreinforced, a positive attitude toward the attitude objectwill form. When behavior toward an attitude object ispunished or associated with negative consequences, anunfavorable attitude will form.

Genetic determinants of attitudes. Identical twins (evenwhen raised in separate environments) show a highercorrelation in their attitudes than fraternal twins, provid-ing evidence for a genetic basis of attitudes. This islikely because of the influence of genetics on tempera-ment and personality, which in turn influence attitudes.Attitudes that have a genetic basis appear to be more dif-ficult to alter and exert a stronger influence on behavior.

IMPLICIT AND EXPLICIT ATTITUDES

A distinction has been made between implicit andexplicit attitudes. An explicit attitude is one that a

person is consciously aware of and can report, forexample, on a self-report measure. A large volume ofresearch has focused on understanding and assessingexplicit attitudes. However, recent attention has turnedto the existence of implicit attitudes, attitudes that areinvoluntary, uncontrollable, and, in some cases, notaccessible at a conscious level. Although implicit atti-tudes are not consciously accessed, they are found tostill exert influence on behavior. Take, for example, aperson who holds sexist attitudes in the workplace butis not consciously aware of holding these attitudes.These are implicit attitudes, which could exert influ-ence on this person’s workplace behavior with regardto female employees. The relationship betweenimplicit and explicit attitudes, along with their influ-ence on behavior, is a topic of ongoing investigationamong attitude researchers. With regard to attitudemeasurement, implicit and explicit attitudes mayrequire different methods of assessment. Becausepeople are not able to directly access and reportimplicit attitudes, traditional means of attitude mea-surement may be less effective, indicating a need formore indirect methods of assessment.

ATTITUDE MEASUREMENT

Attitudes are frequently assessed through self-reportmeasures. Three common scale methodologies used toassess attitudes are the Thurstone scale, Likert scale,and semantic differential. A Thurstone scale is devel-oped by having individuals rank order opinion state-ments about a particular attitude object according totheir favorableness. A subset of items representing awide range of opinions is then selected and used toassess attitudes. A Likert scale consists of a series ofitems for which people indicate the strength of theiragreement with each statement (e.g., “I enjoy my job”)on a rating scale that encompasses low to high levels ofagreement. The semantic differential assesses attitudesby providing opposing adjective pairs (e.g., good–bad;foolish–wise) on which the individual rates a specificattitude object.

Although there are advantages to measuring atti-tudes through direct self-report measures, such asavailability, speed, and ease of use, there are also lim-itations associated with their use. For example, manyexisting self-report measures make the implicitassumption that attitudes are bipolar (rather thanbivariate) and, therefore, may not detect levels of atti-tudinal ambivalence. Further, when individuals are

38———Attitudes and Beliefs


asked to report attitudes on controversial topics, theymay be less likely to report their true evaluations andinstead report responses that they perceive to besocially desirable. Similarly, if attitudes are not con-sciously accessible, as in the case of implicit attitudes,individuals may not be able to accurately report themon these measures. To overcome these concerns,researchers can use indirect methods of attitude mea-surement, such as unobtrusive behavioral measures,physiological measures, or techniques, such as theImplicit Association Test, that are designed for assess-ing implicit attitudes.

DO ATTITUDES PREDICT BEHAVIORS?

The question of whether attitudes guide and predictbehavior is an issue that has been central to the studyof attitudes. Several critical challenges to the com-monsense assumption that attitudes determine behav-ior emerged in the 1930s and 1940s, as numerousstudies indicated little or no relationship between atti-tudes and behaviors. Consequently, by the 1960s therewas a call by many researchers to abandon the studyof the attitude. Since then, researchers have reexam-ined the attitude–behavior link and articulated partic-ular conditions under which attitudes will be likely toguide behavior. Attitudes that are accessible, specific,strong, or formed through direct experience are foundto exert stronger influences on behavior. Additionally,the theory of reasoned action, developed by IcekAjzen and Martin Fishbein, and the theory of plannedbehavior, developed by Ajzen, provide models of howattitudes can guide deliberative behavior through theirinfluence on intentions.

PERSUASION

Persuasion refers to an active attempt made to changeanother person’s attitude toward some issue, object, orperson. Seminal studies conducted during the 1940s byCarl Hovland and his research group atYale Universityled to the development of the message learningapproach, which became a primary template for per-suasion research. The message learning approach sug-gests that persuasion occurs through a sequence ofstages including attention, comprehension, yielding,and retention of a message. It asserts that persuasion isinfluenced by characteristics related to the source ofthe message, the nature of the audience (or messagerecipients), and qualities of the message itself.

In the 1980s, dual-process models, such as ShellyChaiken’s heuristic-systematic model and the elabora-tion likelihood model, developed by Richard Pettyand John Cacioppo, emerged as dominant models ofpersuasion. These models suggest that persuasion canresult from two types of message processing: thought-ful processing of the arguments contained in a mes-sage, or less effortful processing of cues or heuristicspertaining to the message. Whether one engages inmore or less effortful processing depends on one’sability or motivation to elaborate on the message.Although attitude change can occur through eitherprocess, persuasion that results from more elaborativeprocessing of a message has been found to be morepersistent, resistant to counterpersuasion, and predic-tive of future behavior.

—Jennifer L. Welbourne

See also Measurement Scales; Theory of ReasonedAction/Theory of Planned Behavior

FURTHER READING

Cacioppo, J., Gardner, W., & Berntson, G. (1997). Beyondbipolar conceptualizations and measures: The case ofattitudes and evaluative space. Personality and SocialPsychology Review, 1, 3–25.

Kraus, S. (1995). Attitudes and the prediction of behavior:A meta-analysis of the empirical literature. Personalityand Social Psychology Bulletin, 21, 58–75.

Olson, J., Vernon, P., Harris, J., & Lang, K. (2001). Theheritability of attitudes: A study of twins. Journal ofPersonality and Social Psychology, 80(6), 845–860.

Petty, R. E., & Wegener, D. (1999). The elaboration likeli-hood model: Current status and controversies. InS. Chaiken & Y. Trope (Eds.), Dual-process theoriesin social psychology (pp. 41–72). New York: GuilfordPress.

Wood, W. (2000). Attitude change: Persuasion and socialinfluence. Annual Review of Psychology, 51, 539–570.

ATTRACTION–SELECTION–ATTRITION MODEL

The discipline of organizational behavior focuseson the study of organizations and the people whopopulate them. Generally and historically, the fieldhas been largely divided into those who study theattributes of organizations and their markets (macro

Attraction–Selection–Attrition Model———39


organizational behavior) and those who study theattributes of people in organizations (micro organiza-tional behavior). Typically, macro approaches havefocused on explaining organizational performanceand draw their intellectual heritage from sociologyand economics, whereas micro approaches havefocused on explaining and predicting individualbehavior and performance and draw their heritagefrom psychology. Although the recent history of orga-nizational behavior has seen attempts to integratethese two paradigms, the micro and macro distinctionhas led to a scholarly division with two largely non-overlapping, independent literatures. As a conse-quence, there is little cross-fertilization of ideas acrossmicro and macro perspectives and little attempt tounderstand the processes that translate the characteris-tics and behavior of people to the performance of theirorganizations. In his 1985 presidential address to theSociety of Industrial and Organizational Psychology,Benjamin Schneider noted this distinction in the liter-ature and offered an attempt to bridge the micro andmacro distinctions. In its most basic form, his model,the attraction–selection–attrition (or ASA) model,postulates that it is the characteristics of people in anorganization that partially (if not largely) determinethe organizational attributes typically studied by macroresearchers.

OVERVIEW

The ASA model delineates a framework for under-standing organizational behavior that integrates bothindividual (micro) and organizational (macro) perspec-tives by explaining macro organizational attributeswith micro person characteristics. The framework pro-poses that the outcome of three interrelated dynamicprocesses, attraction–selection–attrition, determinesthe kinds of people in an organization, which conse-quently defines an organization, its structures, itsprocesses, and, ultimately, its culture.

At the core of the ASA model are the goals of theorganization originally articulated (implicitly orexplicitly) by the founder. Organizational goals, andthe processes, structures, and culture that emerge tofacilitate attainment of these goals, are suggested to bereflections of the particular characteristics (i.e., per-sonality) of the founder and those of his or her earlycolleagues. Schneider suggests that founders are facedwith a variety of decisions to make regarding whom tohire, how to compensate employees, how to structure

reporting relationships, and even what industries ormarkets to enter. The decisions made are influencedby the underlying values, motives, and dispositions ofthe founder. So, for example, the ASA model wouldpostulate that the cultural differences between AppleComputer and Microsoft had their origins in the per-sonality differences of their founders, Steve Jobs andBill Gates. As Apple Computer and Microsoft grew,the policies and procedures established were a reflec-tion of their founders’ early influence, and over timethese policies and procedures created a culture that issomewhat unique for each company. So, the genesisof an organization’s culture can be traced to the initialdecisions made by founders and the unique imprintthey put on their organizations. This, too, is the begin-ning of the ASA cycle.

The ASA cycle begins with the attraction process,which concerns the fact that people’s preferences forparticular organizations are based on some estimate ofthe fit or congruence of their own personal character-istics (personality, values, and motives) with theattributes of the organization they are evaluating. Thatis, people find organizations differentially attractive asa function of their implicit judgments of the congru-ence between those organizations’ goals (and struc-tures, processes, and culture as manifestations ofthose goals) and their own personalities. For example,an IT engineer may choose to work for AppleComputer, as opposed to Microsoft, because she or hesees the company as innovative and flexible, whichconforms to the engineer’s own values of creativityand independence. Ample research evidence suggeststhat job applicants make assessments of fit whenchoosing among employment alternatives.

The next step in the ASA cycle refers to the formaland informal selection procedures used by organiza-tions in the recruitment and hiring of people with theattributes the organization desires. Many organiza-tions explicitly use fit as a criterion in the hiringprocess. Based on ample research demonstrating thatfit to an organization’s culture has implications foremployee job satisfaction, turnover, and absenteeism,this criterion seems justified. The greater the degree ofmisfit, the more likely an employee will be to experi-ence dissatisfaction with the job, be absent, and quit.Research also suggests that fit assessments affect hir-ing procedures not intended to assess fit. For example,research suggests that assessment center ratings andinterviewer judgments are influenced by conscious orunconscious evaluations of applicant fit.

40———Attraction–Selection–Attrition Model


Finally, the attrition process refers to the idea thatpeople will leave an organization they do not fit. Theturnover literature is quite clear about the fact thatpeople who do not fit an organization will tend toleave it. Of course, economics and job marketprospects moderate the extent to which people leavean organization they do not fit. In summary, ASA pro-poses that three processes—attraction, selection, andattrition—result in organizations containing peoplewith distinct personalities, and it is these distinctpersonalities that are responsible for the unique struc-tures, processes, and cultures that characterize organi-zations. Organizational and personal characteristicsare self-reinforcing. The characteristics of people inan organization determine the policies and practices,which, in turn, determine the people who are attractedto and remain with the organization.

IMPLICATIONS FOR ORGANIZATIONS

As an outcome of the ASA model, Schneider andcolleagues postulated that organizations will becomeincreasingly homogeneous over time. In other words,they will come to be populated by people of a similarpersonality profile.To assess this hypothesis, Schneiderand colleagues examined the personality profiles ofapproximately 13,000 managers from 142 organiza-tions in the United States. The organizations in theirsample represented a broad cross section of industries.Consistent with the homogeneity hypothesis, theirresults suggested that managers were more similar tomanagers in their own organization than they were tomanagers in the other organizations. This remained truewhen you looked within an industry. That is, evenwithin an industry, managers were more similar to oth-ers in their organization than they were to managers inother organizations within their same industry.

Although we previously indicated that there arepositive consequences of good fit for people and orga-nizations (regarding satisfaction, commitment, andturnover), the ASA model suggests that the outcomegood fit could be detrimental to the long-term viabil-ity of an organization, particularly if an organizationexperiences volatility in its market. The primary neg-ative consequences of good fit or homogeneity arethe potential inability for an organization to sensechanges in its environment and adapt to those changesand the demise of competitiveness through easily pre-dictable decision making. There is limited research onthe consequences of homogeneity for organizational

effectiveness, and the predictions made by the ASAmodel are complex. For example, the ASA modelwould predict that during the initial founding andearly history of an organization, homogeneity breedsthe commitment that is needed to retain people andgrow the enterprise. Only after an organization maturesand the market becomes more complex and turbulentdoes homogeneity produce negative consequences.Research does indicate that as the average tenure ofthe senior managers increases, the fit between theorganization’s strategy and demands of the businessenvironment decreases. Although not a direct test ofthe negative consequences of homogeneity, this find-ing is consistent with the logic of the hypothesis.Additionally, research in social psychology on theeffects of homogeneity on group problem solvingsupports the notion that groups engaged in creativeproblem-solving tasks do better if they are heteroge-neous. The conflict that is created by different perspec-tives is important in these ill-defined problem-solvingsituations—situations analogous to the strategic deci-sions made by top managers. As this research implies,the ASA model predicts that the negative conse-quences of homogeneity may only manifest themselvesat the upper levels of the organizational hierarchy(where managers are faced with strategic decisions).Elsewhere the positive benefits of homogeneity mayoutweigh the costs.

IMPLICATIONS FORORGANIZATIONAL BEHAVIOR THEORY

The ASA model provides an example of multilevelorganization theory. Specifically, the psychologicalattributes of people (in the collective) are hypothe-sized to be the antecedents of important organiza-tional characteristics. In this way, the ASA modeloffers a bridge between the micro and macro perspec-tives. Additionally, the ASA model provides insightinto a long-standing argument within psychology—the person–situation debate. This debate seeks todetermine which set of attributes (those related to theperson, or the situation/environment) are the primarypredictors of behavior. The ASA model suggests thatthe attributes of people shape their environments. Thetwo sets of attributes are not mutually exclusive;rather, they are mutually determined. You cannotseparate people from the situation.

—D. Brent Smith

Attraction–Selection–Attrition Model———41


See also Employee Selection; Organizational Develop-ment; Person–Environment Fit; Prescreening Assess-ment Methods for Personnel Selection; Recruitment;Selection Strategies

FURTHER READING

Giberson, T. R., Resick, C. J., & Dickson, M. W. (2005).Embedding leader characteristics: An examination ofhomogeneity of personality and value in organizations.Journal of Applied Psychology, 90, 1002–1010.

Schneider, B. (1983). Interactional psychology and organi-zational behavior. In L. L. Cummings & B. M. Staw(Eds.), Research in organizational behavior (Vol. 5,pp. 1–31). Greenwich, CT: JAI Press.

Schneider, B. (1987). The people make the place. PersonnelPsychology, 40, 437–454.

Schneider, B., Goldstein, H., & Smith, D. B. (1995). TheASA framework: An update. Personnel Psychology, 48,747–773.

Schneider, B., Smith, D. B., Taylor, S., & Fleenor, J. (1998).Personality and organization: A test of the homogeneityof personality hypothesis. Journal of Applied Psy-chology, 83, 462–470.

AUTOMATION/ADVANCEDMANUFACTURINGTECHNOLOGY/COMPUTER-BASEDINTEGRATED TECHNOLOGY

Automation usually refers to the replacement of humanwork by machines. The word was first used by the FordMotor Company in the 1940s to describe automatichandling and machine-feeding devices in their manu-facturing processes. Advanced manufacturing technol-ogy (AMT) is a special instance of automation andusually refers to computer-based manufacturing tech-nologies and support systems. Examples includecomputerized numerically controlled machine tools,computer-aided design, and computer-supported pro-duction control systems. There will be few, if any, man-ufacturing companies in the developed world that havenot undertaken some investment in AMT.

Computer-based integrated technology (CIT)refers to higher levels of integration and comprisessystems that cut across organizational functions. Forexample, enterprise resource planning (ERP) systemsinclude a centralized database and sets of integratedsoftware modules designed to manage all aspects of

an organization’s work processes, including produc-tion control, customer billing, and human resources.Estimating the uptake of CIT is difficult. However, asurvey in Australia, Japan, and the United Kingdom,published in 2002, found that approximately 33% to40% of larger manufacturing companies (employingmore than 250 people) were significant users of CIT.The same survey in Switzerland reported substantialuse in around 60% of companies. The findings aresimilar for ERP systems. By the late 1990s, it wasestimated that around 40% of large U.S. companiesand 60% of small ones had deployed ERP systems. By2004, the worldwide market for ERP systems wasestimated to be around $79 billion per annum.

Over the last decade, there has also been growinginvestment in systems to integrate activities betweenorganizations, a good example being e-businesssystems that allow electronic ordering and billingthrough a supply chain and on the part of customers.By the year 2000 it was estimated that around 20%to 25% of companies in the United States, Canada,Europe, and Australia were trading online, althoughthe proportional value of goods traded online wasmuch lower (less than 10%). It is almost certainly thecase that these amounts have grown and will continueto grow.

MOTIVES AND IMPACTS

Such investments are usually undertaken for a mix ofmotives. Machines may do the work more cheaply,more quickly, to a higher quality, with more repeata-bility, with reduced errors, and with reduced leadtimes. For these reasons, many companies havebecome enthusiastic adopters of such new technolo-gies. They are also mindful that if they don’t innovate,their competitors might, thereby gaining a significantadvantage in the marketplace. This can feed so-calledfads and fashions, often vigorously supported byan active community of suppliers of equipment andexpertise, including consultants.

Unsurprisingly, such changes are also often accom-panied by fears on the part of employees. Will theadoption of new technology lead to reduced head-count and thereby redundancy? Will the remainingjobs become deskilled, with previously skilledemployees being reduced to unskilled labor?

It is certainly the case that the trend to automationcan reduce headcount. To give a specific example, thecity of Sheffield in the United Kingdom, famous for

42———Automation/Advanced Manufacturing Technology/Computer-Based Integrated Technology


its high-quality steel, produces the same amount as itever did in its postwar prime, but now with 10% of theearlier workforce.

But at the same time, the development of comput-ers and their increasing application to differentdomains has spawned whole new industries, therebycreating many new jobs. New organizations havegrown up around the development, provision, market-ing, and support of computer hardware and software,project management, knowledge management, com-puter simulations, software games and entertainment,and communications, to name just some—all enabledby the onset of sophisticated computerization.

Concerns over deskilling are equally complicatedto assess in practice. Whereas some organizationshave used computer-based operations to deskill theiroperators—for example, by turning them intomachine minders—many others have upskilled theiroperations by asking their machine operators to writeand edit computer programs and to solve complexmachine problems. Also, as previously implied, ata more macro level, the onset of computerizationhas led to the creation of many new highly skilledprofessions.

The process is further complicated by the onset ofglobalization. Computer-based information and com-munications technologies now make it possible tomove work around the world. A topical example isprovided by the widespread use of customer call cen-ters based in India. This may be to the benefit of theIndian economy, but it may not be perceived that wayby employees in the developed world who see theirjobs as being exported to regions where labor costs aresignificantly lower.

Three generalizations seem appropriate. First, suchperiods of change may be genuinely uncomfortableand threatening for the individuals concerned. It maybe no real consolation in losing one’s job to be told itis an inevitable long-term structural shift in the natureof the global economy. Second, such changes arelikely to be easier to manage and endure during peri-ods of economic growth rather than decline. A buoy-ant labor market certainly helps. And third, this is oneof the reasons why most leading commentators indeveloped economies see their economic future in thedevelopment of highly skilled, high value-added, andhighly innovative work, areas where education andskills are at a premium and where competition in aglobal economy is not solely dependent on the costof labor.

EFFECTIVENESS AND THE ROLE OFINDUSTRIAL/ORGANIZATIONAL PSYCHOLOGY

The foregoing description gives the impression ofinevitability and, although difficult perhaps for someindividuals in the short term, of benign and effectiveprogress. However, the position on the ground is a gooddeal more complex. Let us look now at some of the dataon the effectiveness of such new technologies.

The data from economic analyses, surveys, casestudies, and expert panels is consistently disappoint-ing. Turning first to ERP systems, many are scrapped(estimates vary between 20% and 50%), and overallfailure rates are high (again, estimates vary, at around60% to 90%). Indeed, it is now commonplace foreconomists to bemoan the lack of impact of invest-ments in IT (information technology) on overall pro-ductivity over time.

The best estimate is probably that up to 20% ofinvestments are true successes, genuinely meetingtheir goals; around 40% are partial successes, meetingsome of their goals but by no means all; and around40% are complete failures.

So, why are such investments often so disappoint-ing, and what can be done about it? Many indus-trial/organizational psychologists have worked in thisdomain, most notably perhaps under the general ban-ner of sociotechnical thinking. Their central proposi-tion is that work systems comprise both technical andsocial systems and that companies cannot change onewithout affecting the other—it is the nature of systemsthat they are intrinsically interconnected. It followsthat technical change requires active considerationto changes in working practices and processes, jobdesigns and work organization, employee skills andcompetencies, training and education, human–computer interfaces, and the management of change.These are major issues, and the evidence is that manyorganizations focus too much on the technology, paytoo little regard to the social, and fail to adopt an inte-grated systems perspective.

Several attempts have been made at formulatinggood practice guidelines, of which the following arerepresentative:

• Senior managers should ensure that new technologyinvestments meet the needs of the business. Seniormanagers should ask, “Why are we doing this? Whatbenefit do we gain? Does it further our strategy?”

• Any technical change will require changes in busi-ness processes, working practices, job design, and

Automation/Advanced Manufacturing Technology/Computer-Based Integrated Technology———43


the like. Senior managers need to ensure that changesin all these areas are an intrinsic part of the project—a systems view is needed.

• Senior users in the business need to have someownership of, and influence over, the nature of thechanges they require. Changes in systems that arepulled into a business are usually much more suc-cessful than changes that are pushed into a business.Beware projects that seem just to be about IT andthat are being pushed hard by the IT department.

• Any project team needs to include all the requisiteskills and expertise, including the human and organi-zational issues.

• The users (or recipients) of any change program needto be actively involved. This should be all the wayfrom the design of the new way of working throughto evaluation of the effectiveness of the changes.

• There is a need to educate all those involved in whatthe changes mean, why they are being undertaken,what benefits accrue, and what actions are necessaryto achieve success. At the same time, training isneeded on the operational and more detailed aspectsof the changes.

• Where such changes are undertaken, organizationsneed to learn as they go, to be pragmatic, and, wherepossible, to undertake changes in manageablechunks.

• Evaluation against objectives using benchmark mea-sures is a prerequisite for learning. Internal andexternal benchmarking can provide excellent oppor-tunities for improvement.

• All the above require the commitment of resources,in particular time, effort, money, and expertise. Theyalso require a different mind-set on the nature ofchange, one that adopts a systems orientation andviews technology as a necessary but not sufficientpredictor of success.

These guidelines may seem relatively unsurprisingto students of industrial/organizational psychology.

But there continues to be evidence that such standardsare ignored in practice. Perhaps the interesting ques-tion is, “Why is it that informational technologyfailures persist?” There is massive potential here forindustrial/organizational psychologists to make a sub-stantial contribution, but it is likely that this will bestbe achieved by working with other disciplines (includ-ing technical and operational specialists) and withorganizations facing some very practical problems. Itis certainly true that we need to bridge the dividesbetween disciplines and between academia andpractice.

—Chris Clegg

See also Computer Assessment; Human–ComputerInteraction; Simulation, Computer Approach

FURTHER READING

Clegg, C. W. (2001). Sociotechncial principles for systemdesign. Applied Ergonomics, 31, 463–477.

Clegg, C. W., Wall, T. D., Pepper, K., Stride, C., Woods, D.,Morrison, D., et al. (2002). An international survey ofthe use and effectiveness of modern manufacturingpractices. Human Factors and Ergonomics in Manufac-turing, 12, 171–191.

Holman, D., Wall, T. D., Clegg, C. W., Sparrow, P., &Howard, A. (2003). The new workplace. Chichester,UK: Wiley.

Landauer, T. (1995). The trouble with computers. Cambridge:MIT Press.

AUTONOMY

See EMPOWERMENT

44———Automation/Advanced Manufacturing Technology/Computer-Based Integrated Technology


ABUSIVE SUPERVISION - Sage Publications

Documents