Murphy & Kram 2010

8/10/2019 Murphy & Kram 2010

1/27

Understanding non-workrelationships in developmental

networksWendy Marcinkus Murphy

Department of Management, College of Business, Northern Illinois University,DeKalb, Illinois, USA, and

Kathy E. KramDepartment of Organizational Behavior, School of Management,

Boston University, Boston, Massachusetts, USA

Abstract

Purpose The purpose of this study is to explore the different contributions of work and non-workrelationships that comprise individuals developmental networks to career success.

Design/methodology/approach A multi-method approach provides a rich understanding of howwork and non-work developmental relationships combine to support individuals careers. Survey datawere analyzed from 254 working adults who were also part-time MBA students. Semi-structuredinterviews were conducted with 37 participants.

Findings Quantitative results indicate that non-work developers provide more overall supportthan work developers. Support from non-work developers is positively associated with careersatisfaction and life satisfaction. In contrast, support from work developers is positively associatedwith salary level and career satisfaction. Qualitative data indicate differences in the sub-functionsand quality of support offered by work versus non-work relationships, particularly in terms of rolemodeling.

Research limitations/implications Developmental relationships from different domainsemphasize different sub-functions of support and differentially affect career outcomes. While broadfunctions career support, psychosocial support, and role modeling are identifiable across domains,non-work relationships provide some distinct sub-functions from work relationships.

Practical implications Practicing managers should develop and maintain developmentalnetworks that extend beyond the boundaries of their current organization. Human resourceprofessionals will want to consider how well their initiatives encourage individuals to enlist a varietyof potential developers into their networks.

Originality/value The findings indicate that non-work relationships are a critical part ofdevelopmental networks and individuals career success.

KeywordsEmployee relations, Interpersonal relations, Career development, Part time students,

Social networks, United States of AmericaPaper typeResearch paper

The current issue and full text archive of this journal is available at

www.emeraldinsight.com/1362-0436.htm

An earlier version of this paper was presented at the Annual Meeting of the Academy ofManagement in Annaheim, California, August 2008. The authors appreciate helpful commentsfrom Dawn Chandler and Monica Higgins. The first author thanks the American Association ofUniversity Women for their generous support of this research through a dissertation fellowship.

Understandingnon-work

relationships

637

Received 28 April 2010Revised 5 August 2010

Accepted 18 August 2010

Career Development InternationalVol. 15 No. 7, 2010

pp. 637-663q Emerald Group Publishing Limited

1362-0436DOI 10.1108/13620431011094069

8/10/2019 Murphy & Kram 2010

2/27

IntroductionDevelopmental relationships have been a topic of interest for both scholars andpractitioners for 25 years. Interest in this topic persists because relationships withmentors may be especially crucial in the turbulent, changing career environment of the

twenty-first century (Allen et al., 2004; DeLong et al., 2008; Kram, 1996). Recentmentoring research has broadened the inquiry to include multiple relationships thatcomprise ones developmental network as the source of support for an individualscareer (Higgins and Kram, 2001; Molloy, 2005). This study extends our understandingof how work and non-work relationships that individuals include in theirdevelopmental networks differentially affect important career outcomes. We use amixed method approach to examine support provided by developmental networkrelationships both within and beyond the context of work organizations, extendingprevious research on alternative sources of developmental support (Allen andFinkelstein, 2003; Eby, 1997) and external mentors (Baugh and Fagensen-Eland, 2005).In addition, we answer the calls for empirical work linking developmental relationshipsto outcomes that bridge the work-life interface, such as life satisfaction (Dougherty and

Dreher, 2007; Greenhaus and Singh, 2007).Several scholars suggest that developmental relationships are likely to exist both

within and outside work organizations (Baugh and Fagenson-Eland, 2005; Higgins andKram, 2001). Research on developmental networks has demonstrated that individualsreceive support from relationships in both their professional and personal social arenasamong others (Higgins, 2000; Higgins and Thomas, 2001; Cummings and Higgins,2006). Empirical studies indicate that mentors most often provide assistance in theforms of career (or vocational) and psychosocial support (Allen et al., 2004; Kram,1985). Similarly, developers who comprise an individuals developmental networkcombine to provide both career and psychosocial support (Higgins, 2000; Higgins andKram, 2001)[1]. Recent developmental network studies suggest that networkcomposition and structure will shape which career outcomes are realized (Dobrowand Higgins, 2005; van Emmerik, 2004). Despite this accumulated knowledge, we stilllack an empirical understanding of how work developers and non-work developers inan individuals network may offer distinctive types of support and differentially affectcareer success. Formally, the research questions that define this study are:

RQ1.What are the contributions of both work and non-work developmentalrelationships to individual career outcomes, including salary level, careersatisfaction, and life satisfaction?

RQ2.How do support functions from non-work relationships differ from workrelationships in developmental networks?

First, we review the conceptual underpinnings that guided our mixed method empiricalstudy. We then present quantitative data that links both work and non-workdevelopmental support to important career outcomes, and qualitative data thathighlight the most salient functions provided by developers from different domains.Our findings indicate that non-work relationships are a critical part of developmentalnetworks, offering unique support functions that contribute to individuals careersuccess. We end with suggestions for future research in this arena, and some practicalimplications for individuals and human resources professionals as they strive to fosterlearning and career development in organizations.

CDI15,7

638

8/10/2019 Murphy & Kram 2010

3/27

Developmental networksDevelopmental networks are conceptualized as a set of relationships that offermentoring assistance including senior colleagues, peers, family, and communitymembers (Higgins and Kram, 2001; Molloy, 2005). A developmental network is an

ego-centered network; its structure includes all of the relationships an individualidentifies in which the other takes a special interest in his or her development andcareer-related issues are discussed. Developmental relationships are a type of highquality connection with theoretical roots in adult development, social exchange theory,social networks, and identity (Dutton and Heaphy, 2003; Levinson, 1978; Ragins andKram, 2007). High quality connections allow co-construction of identities or relationalidentification, which in the context of careers suggests that support from developersenable a positive evaluation of ones work role, performance, and well-being (Sluss andAshforth, 2007).

The recent emergence of research on positive organizational psychology (Cameronet al., 2003) has called attention to the critical role that work relationships, particularlydevelopmental relationships, have in producing positive individual and organizationaloutcomes (Dutton and Heaphy, 2003; Dutton and Ragins, 2007). The few studies usinga network approach indicate that an individuals developmental network explainsgreater variance in career outcomes than simply examining traditional mentoringrelationships, the supervisor-subordinate relationship, or coworker relationships(Higgins and Thomas, 2001; Ibarra, 1992, 1995). Yet, to date, there has been very littleattention to how non-work developmental relationships may contribute to positivecareer outcomes. Here, we review what we know thus far about the contributions ofboth work and non-work developmental relationships to valued individual outcomes.

Work developmental relationshipsExtensive research shows that mentors at work provide career and psychosocial

support that enhances individuals career outcomes (e.g. Kram, 1985; Noe, 1988;Scandura, 1992); however, recent research suggests that a network of developers couldcombine to provide such support (Higgins and Kram, 2001). According to Kram (1985),career functions enhance career advancement while psychosocial functions enhanceproteges sense of competence and effectiveness in a professional role. In ameta-analysis of traditional mentor-protegerelationships, Allenet al.(2004) found thatcareer and psychosocial support were related to both salary level and careersatisfaction. These results indicated that while career support was more stronglyassociated with objective career outcomes, the effect sizes were relatively small (Allenet al., 2004). Although Scandura (1992) found empirical support for treating rolemodeling as a distinct function of mentoring relationships, few studies have testedfunctions beyond the two core dimensions of career support and psychosocial support

(Allen et al., 2004). Studies also show that informal mentoring relationships providemore support and have a stronger influence on career outcomes than formal mentoringrelationships (Raabe and Beehr, 2003; Ragins and Cotton, 1999).

Expanding the scope to developmental networks, research findings indicate that themore people an individual identifies as being in his or her developmental network(greater size), the greater his or her work satisfaction (Higgins, 2000). Similarly, greatersize, emotional intensity, and frequency of contact with ones developmental networkare associated with greater intrinsic career success (van Emmerik, 2004). In addition,

Understandingnon-work

relationships

639

8/10/2019 Murphy & Kram 2010

4/27

Higgins and Thomas (2001) found that the developmental network explained morevariance in work satisfaction than when only the traditional primary developer wasassessed. Research has confirmed that the social system developers come from matters(Baugh and Fagenson-Eland, 2005; Seibert et al., 2001); in other words, the domain in

which relationships originate affects their impact. As social resources, developers inthe work domain may provide workplace focused support and benefits, such assponsorship or protection, which non-work developers are unable to provide (Baughand Fagensen-Eland, 2005). Thus, we expect that support from developers in the workdomain primarily affect career outcomes in the work domain. Specifically, wehypothesize that:

H1a. Support from work developers will be positively related to salary level.

H1b. Support from work developers will be positively related to career satisfaction.

Non-work developmental relationships

Scholars advocate a multiple mentor approach for individual growth (de Janasz et al.,2003; Higgins and Kram, 2001) and argue that developmental relationships are acritical component of career development and success (Baugh and Sullivan, 2005).Allen and Finkelstein (2003) report that 75 percent of their participants (n 88) had atleast one alternative source of developmental support, defined as support outside oftraditional mentoring relationships, and encourage future researchers to examine theefficacy of various sources of support as they relate to objective and subjectiveindicators of career success. Yet few studies on non-traditional mentoring ordevelopmental networks have examined non-work relationships (for exception, seeCummings and Higgins, 2006). Instead, researchers have specified only workrelationships (Baugh and Fagensen-Eland, 2005; de Janasz and Sullivan, 2004; vanEmmerik, 2004) or combined work and non-work domains in analyses (Higgins, 2000,

2001; Higgins and Thomas, 2001; Dobrow and Higgins, 2005). Non-work relationshipsas they relate to careers have been primarily explored in the social support andwork-family literatures, although the focus in these areas tends to be on health issues(e.g. stress and burnout) and work-family conflict/facilitation respectively. Theseliteratures however, offer insights for considering how non-work developmentalrelationships function to affect career outcomes.

Social support has been most widely studied in the stress literature as both anantecedent and moderator of the stress-strain relationship (Carlson and Perrewe, 1999).Studies generally find that within domain effects are stronger than cross-domaineffects (Cutrona and Russell, 1990), for example work support is linked to jobsatisfaction and spousal support is linked to family satisfaction (Parasuraman et al.,1992). However, non-work relationships, particularly those with parents, influence

career and life aspirations (Lent and Brown, 1996), against which subjective success isassessed by the individual (Hall, 2002; Heslin, 2005). In addition, researchers havelinked the existence of relationships and resources in the non-work domain to workoutcomes including organizational commitment (Cohen, 1995; Cohen and Kirchmeyer,1995), salary (Landau and Arthur, 1992), and turnover (Leeet al., 2004).

Recent work-family research also offers insights on the contributions of non-workrelationships to careers. Greenhaus and Powell (2006) suggest that work and familydomains may enrich one another if the experiences in one role improve the quality of

CDI15,7

640

8/10/2019 Murphy & Kram 2010

5/27

life in the other role. Positive (and negative) attitudes may spillover from one domain tothe other or crossover from spouse to spouse (Kirchmeyer, 1992, 1993; Rothbard, 2001).Studies have linked marriage and support from family to work-family conflict orfacilitation, which in turn affects career outcomes including income and job satisfaction

(Adamset al., 1996; Froneet al., 1997; Landau and Arthur, 1992). Overall, such workoutcomes as a consequence of the interaction between work and family are more highlyrepresented in the work-family literature than non-work outcomes (Eby et al., 2005).Nonetheless, research generally suggests that non-work relationships may positivelyinfluence outcomes across both work and non-work domains.

With the rapid pace of change and the modern boundaryless career context,researchers suggest that individuals will develop their careers across multipleorganizations (Arthur and Rousseau, 1996). In this environment, non-workrelationships in developmental networks tend to be more stable than workrelationships (Cummings and Higgins, 2006). Therefore, we argue thatdevelopmental support is an important and overlooked mechanism linking non-workrelationships to outcomes across the work-life interface. When non-work relationshipsare included in ones developmental network, it is likely that the developmentalsupport received from these sources will be positively related to important careeroutcomes, including career satisfaction and salary level, as well more holisticoutcomes, such as life satisfaction. Thus, we hypothesize that:

H2a. Support from non-work developers will be positively related to lifesatisfaction.

H2b. Support from non-work developers will be positively related to careersatisfaction.

H2c. Support from non-work developers will be positively related to salary level.

Developmental networks and career successCareer success is an outcome of interest for both individuals and organizations. Forindividuals, a combination of objective success (e.g. salary, promotion, status) andsubjective success (e.g. self-esteem, psychological success, sense of well-being) arepositive outcomes that may be facilitated by developmental relationships thatencourage conscientious career management (de Janaszet al., 2003; Kram, 1996). Froman organizational perspective, successful individuals with more supportiverelationships are more likely to be satisfied and committed to their jobs and lesslikely to turnover (Allenet al., 2004; Leeet al., 2004; Ragins and Kram, 2007).

Several careers researchers suggest that people conceptualize and evaluate theircareer success in realms that go beyond their work boundaries (e.g. Arthur and

Rousseau, 1996; Hall, 2002; Heslin, 2005). This is important when examining the impactof non-work developmental relationships that are likely to affect outcomes that crossthe work-life interface. Well-being is a particularly relevant outcome since it is anindividuals overall assessment of combined roles, from both work and non-workdomains. As a proxy for subjective well-being, researchers often use global lifesatisfaction, which is conceptually related to happiness (e.g. Judge et al., 2005; Lentet al., 2005). Thus, in this study, subjective career success includes both careersatisfaction and life satisfaction.

Understandingnon-work

relationships

641

8/10/2019 Murphy & Kram 2010

6/27

Support from work versus non-work developersKrams (1985) original qualitative study defined mentoring and its career andpsychosocial functions that facilitate career success. Few studies since have exploreddevelopmental relationships using qualitative inquiry and those that have focus on

traditional, work-based relationships. For example, Eby and Lockwood (2005) foundthat both mentors and proteges in a formal mentoring program reported the positivebenefits identified by Kram (1985); however negative aspects were also reported. Otherstudies have explored developmental relationships from the mentors perspective orfocused on negative experiences (Allenet al., 1997; Ebyet al., 2000). One exception is astudy by Allen and Finkelstein (2003), which confirmed that individuals receive careerand psychosocial support from nontraditional developmental relationships, withcoworkers as the most frequently cited alternative.

We argue that a combined quantitative and exploratory qualitative approach isparticularly important at this juncture in the mentoring and developmental networksliterature, since our knowledge has primarily been developed from studies onwork-based relationships. In this study, we use Pelligrini and Scanduras (2005) scale,which has general items applicable to the kinds of support that both work andnon-work developers could provide. This is because it is unlikely that non-workdevelopers could provide some of the sub-functions of career support originallyidentified by Kram (1985), including exposure and visibility, protection, andchallenging assignments. These sub-functions are also unlikely to be provided byextra-organizational mentors (Baugh and Fagenson-Eland, 2005). Therefore,contrasting the experiences of individuals with work and non-work developmentalrelationships is necessary to identify potential similarities and differences in howsupport is provided from different domains. In addition, data from interviews aid inquantitative data interpretation and enrich our understanding of how developmentalnetworks facilitate career success.

MethodologyThis study uses a multi-method design, which began with the collection of quantitativedata followed by the collection of qualitative data. We conducted a survey in the summerof 2006. The questionnaire included demographic information and a series of previouslyvalidated scales to measure variables of interest. Hypotheses were tested usinghierarchical regression analysis. We conducted 37 follow-up interviews over the winterand spring of 2007 in order to enhance our understanding of the statistical findings.

SampleThe sample includes 254 full-time working professionals, ages 22 to 51, who were alsopart-time MBA students at a major eastern university. Participants were students in

core courses across disciplines, including strategy, organizational behavior, marketing,finance, and operations. The researchers were not instructors for the 285 studentsenrolled, and 95 percent participated in the survey. The 17 participants who were onlyemployed part-time were not included in this analysis. Participation did not affectstudents grade, however they were invited to submit their names for a lottery for $500cash as an incentive. Participants mean age was 29.5 and gender composition was 65.7percent male and 34.3 percent female[2]. Ethnic distribution was 77.2 percentCaucasian/White, 16.1 percent Asian, 3.1 percent Black, 1.6 percent Hispanic/Latino,

CDI15,7

642

8/10/2019 Murphy & Kram 2010

7/27

1.6 percent Mixed ethnicity/racial background, and 0.4 percent did not report ethnicity.The participants worked in a wide range of industries as shown in Table I.

Among survey participants, 119 volunteered for follow-up interviews and 37 werepurposefully sampled (Wimpenny and Gass, 2000) to demographically reflect the

larger sample based on age, gender, and industry of employment. Interviewparticipants mean age was 29.6, with a range of ages from 23 to 48. Gendercomposition was 67.6 percent male and 32.4 percent female. Ethnic distribution was66.7 percent Caucasian/White, 19.4 percent Asian, 5.6 percent Black, 2.8 percentHispanic/Latino, 2.8 percent Mixed ethnicity/racial background, and 2.8 percent did notreport ethnicity. The participants also worked across a wide range of industries, asshown in Table I. Semi-structured interviews lasted from 25 to 45 minutes and weredesigned to probe relationships included in the developmental network and how theserelationships impacted participants perceptions of career success. The interviewingprocess was concluded when we were hearing significant amounts of redundantinformation, additional interviews added no new insights into the phenomenon of

interest (Glaser and Strauss, 1969; Sandelowski, 1995; Strauss and Corbin, 1990), andwe had achieved a representative sub-sample of the larger sample. This approach isconsistent with other qualitative sampling in the extant literature (e.g. Denzin andLincoln, 1994; Ely, 1995; Gordon and Whelan-Berry, 2005; Reayet al., 2006).

Survey participantsInterview

participantsIndustry of employment n % n %

Finance/financial services 88 35.9 14 37.8Accounting 22 9.0 5 13.5Technology/MIS 20 8.2 4 10.8

Engineering 19 7.8 2 5.4Education 14 5.7 1 2.7Manufacturing 10 4.1 2 5.4Banking 9 3.7 1 2.7Consulting 9 3.7 1 2.7Marketing/advertising 9 3.7 1 2.7Biotechnology/pharmaceuticals 8 3.3 1 2.7Health services 8 3.3 1 2.7Othera 6 2.4 0.0Real estate/construction 6 2.4 2 5.4Non-profit 5 2.0 1 2.7Government/public sector 4 1.6 0.0Human resources 4 1.6 0.0Military/defense 4 1.6 0.0

Retail 3 1.2 1 2.7Athletics 2 .8 0.0Energy 2 0.8 0.0Law/legal services 2 0.8

0.0Total 254 103.7 37 100.0

Note: aOther includes one participant each from broadcasting, hospitality, insurance, sales, telecom,travel/tourism

Table I.Participants industry of

employment

Understandingnon-work

relationships

643

8/10/2019 Murphy & Kram 2010

8/27

MeasuresCareer outcomes.Salary was measured by individual gross annual income reported ineight categories: under $15,000; $15,000-24,999; $25,000-$34,999; $45,000-$54,999;$55,000-$74,999; $75,000-$100,000; and over $100,000. Career satisfaction was assessed

using Greenhauset al.(1990) five-item career satisfaction scale with a 0.86 for thisstudy. A sample item is, I am satisfied with the progress I have made toward meetingmy overall career goals. Life satisfaction was measured using Diener et al. (1985)five-item scale with a 0.85 for this study. A sample item is, The conditions of mylife are excellent.

Work and non-work developmental relationships.Developmental network data wascollected with the name generator device developed by Higgins (2000, 2005).Participants were asked to list individuals who have taken an active interest in andaction to advance their career by assisting with their personal and professionaldevelopment, starting with the most influential (Higgins, 2000). They were given sixblank columns to be labeled with first names or initials and asked a series of questionsabout each person. Participants indicated What is the primary social arena that you

know this person from? among the categories of work, family, school, community.The categories of family (n 385), school (n 70), and community (n 45) wereincluded in the non-work social arena since the number of developers from school andcommunity was relatively small. Participants also indicated each persons relationshipto them (e.g. supervisor, coworker, spouse, friend). The total number of individuals thatparticipants listed indicates network size. Work developers included upper managers,supervisors, and coworkers, while non-work developers included spouse/partner,parent, sibling, friend, extended family, neighbor (see Table II).

Developmental support.The nine-item mentoring functions scale by Pelligrini andScandura (2005) measured developmental support. This scale includes three questionsto assess each function, for example, He/she helps me coordinate professional goals(career support), I share personal problems with him/her (psychosocial support),I try to model my behavior after him/her (role modeling). Participants answered these

Social roles of developers n % Mean developmental support Std dev.

Work 387 43.6 3.64 0.68Upper manager 143 16.1 3.52 0.55Supervisor 136 15.3 3.56 0.70Co-worker 70 7.9 3.87 0.55Former supervisor 17 1.9 3.96 0.82Othera 21 2.4 3.93 0.97Non-work 500 56.4 3.92 0.58Parent 213 24.0 3.93 0.55

Spouse/partner 104 11.7 4.07 0.62Friend 74 8.3 4.02 0.41Sibling 39 4.4 3.87 0.48Extended family 33 3.7 3.53 0.29Professor 22 2.5 3.64 0.17Othera 15 1.7 3.68 0.86Total 887 100.0 3.80 0.62

Note: aOthers includes subordinate, client, classmate, neighbor, athletic coach, etc.

Table II.Developer roles and meandevelopmental support

CDI15,7

644

8/10/2019 Murphy & Kram 2010

9/27

questions for each relationship they identified as part of their developmental network,similar to Higgins (2000). For measures of support from work and non-workdevelopers, we took the average amounts of each type of support provided by all therelationships in that domain, consistent with previous analyses of network support

(Higgins, 2000; Higgins and Thomas, 2001).Control variables.Prior research has shown that gender and race may impact an

individuals career success (Greenhaus et al., 1990; Judge et al., 1995). Gender wasdummy coded zero for males and one for females. Ethnicity was coded zero forWhite/Caucasian and one for all others. Salary has been associated with age (Judgeet al., 1995); therefore age was included as a control. Regardless of age, individuals mayhave considerable differences in relevant career experience, thus tenure in career fieldwas included. ANOVA revealed significantly lower mean salary levels for participantsemployed in the Non-profit and Government/Public sector fields compared with allother industries (see Table I), so we include this as a control by dummy coding thesetwo categories 1 and all other industries 0. Organization size was included as a controlto account for the opportunity to participate in informal developmental relationships;generally the larger the organization the more potential work developers available. Toassess organization size, participants were asked, How many people are employed byyour current organization? and selected among six categories: under 100, 101-500,501-1,000, 1,001-5,000, 5,001-9,999, over 10,000. Finally, network size indicates thenumber of developers that participants listed in their developmental network. Thisvariable is included to distinguish the effects of developmental support from thenumber of developmental relationships consistent with previous network analyses(Higgins and Thomas, 2001; Seibert et al., 2001).

Qualitative analysis. Interviews were analyzed using content analysis assisted byHyperResearch software for tracking code creation and counts. As a qualitativemethod, content analysis is used for making replicable and valid inferences from data

to their context, by objectively and systematically identifying specified characteristics(Holsti, 1969). The purpose of this methodology is to look for common ideas andpatterns in participants responses. Our unit of analysis was the sentence or group ofsentences that formed a complete, meaningful thought, and sometimes an entireparagraph could be captured by one code (Miles and Huberman, 1994).

The broad categories of career support, psychosocial support, and role modelingwere used in the first round of coding. In the second round of coding, we used Krams(1985) sub-categories to assess the alignment of developers support with that oftraditional mentoring. In the career support category, these included sponsorship,exposure and visibility, coaching, protection, and challenging assignments. In thepsychosocial support category, these included counseling, acceptance andconfirmation, and friendship. New subcategories were created for the broad role

modeling function, which Kram (1985) had included under psychosocial support, or ifthe data did not fit into any pre-existing codes. In the role-modeling category, four newsubcategory codes were created including career behaviors to emulate, work ethic andvalues, devaluing relationships, and work-life interface failure. The subcategory ofencouragement and emotional support was created under psychosocial support. Thefirst author and an experienced graduate student, who had no prior knowledge of thestudy, coded the interviews separately yielding an inter-rater reliability of 86 percent.Disagreements in coding were resolved through discussion, similar to the process used

Understandingnon-work

relationships

645

8/10/2019 Murphy & Kram 2010

10/27

by Eby and Lockwood (2005). The quotes presented here are representative of thepatterns of relationships and outcomes that are reflected in the 37 interviews.

ResultsDevelopmental network data used for this analysis includes 887 relationships reportedby 242 participants, 12 participants reported no developers. Developmental supportwas reported from both work and non-work relationships for 85 percent ofparticipants. For most participants (58 percent), the primary developer, defined as theperson listed first in the developmental network (Higgins and Thomas, 2001), was afamily member, either a spouse/partner or parent. While 50 percent of secondary ortertiary developers were from work and a vast majority (71 percent) were relationshipswith direct supervisors or upper level managers. Peer relationships accounted for 19percent of all developers, including coworkers, friends, classmates, and neighbors. Thenumber of developers and mean developmental support provided by relationshipsfrom each social arena (work versus non-work) are reported in Table II. From the work

domain, developers primarily included supervisors, upper managers, and coworkers.From the non-work domain, developers primarily included parents, spouse/partner,friends, and others.

The means, standard deviations, and correlations of the variables included in thisstudy are reported in Table III. The correlation between salary and career satisfactionwas low (.18), while salary and life satisfaction were not significantly correlated. Thecorrelation between career satisfaction and life satisfaction was not as low (.40)indicating some overlap. In preliminary analyses, the correlations among supportfunctions (career support, psychosocial support, and role modeling) were all greaterthan 0.80, signaling a potential multicollinearity issue for regression analysis.Regression analysis confirmed a collinearity problem among these functions withtolerance scores below 0.10 (0.05-0.09) and VIF scores greater than 10 (10.7-22.3)

(Allison, 1999). Principal component analysis using Varimax rotation revealed sixfactors accounting for 86 percent of the variance with components distinguishable bydeveloper rather than by developmental support function as expected. Thus, wecreated variables to assess average overall developmental support from work andnon-work domains. Results indicate a significant difference in the average overallsupport received from work developers (M 3.64) compared with non-workdevelopers (M 3.92), t(254) 2 5.54,p , 0.001.

The results for hypothesis testing are shown in Table IV. Several control variablesincluding age, ethnicity, tenure in field, and organization size were significantlyassociated with salary. Age was also significantly associated with both careersatisfaction and life satisfaction. In addition, developmental network size waspositively associated with life satisfaction. Support from work developers is positively

related to bothsalary(b 0.14, p , 0.01)andcareersatisfaction(b 0.18, p , 0.01),supportingH1a and H1b. Support from non-work developers is positively associatedwith both l ife satisfaction b 0:15;p , 0:05 and career satisfactionb 0:11;p , 0:05, supporting H2a and H2b. However, support from non-workdevelopers is not significantly associated with salary, thus,H2cis not supported.

In exploratory post-hoc analysis, we examined the contributions of specificdevelopmental relationships to all three outcome variables. Specific relationships wereincluded if the total number of developers was reasonably high enough for regression

CDI15,7

646

8/10/2019 Murphy & Kram 2010

11/27

Variables

Mean

SD

1

2

3

4

5

6

7

8

9

10

11

1Gender

0.34

0.48

2Age

29.48

4.9220.14*

3Ethnicity

0.77

0.4220.04

20.03

4Tenureincareer

5.94

4.0320.11

0.68***

0

.06

5Industry

0.04

0.19

0.09

0.03

0

.05

20.12

6Organizationsize

4.28

1.8720.07

20.01

20

.18**

0.01

20.16*

7Networksize

3.57

1.65

0.13*

20.09

0

.13*

0.05

0.01

2

0.11

8Salary

5.20

1.6520.15**

0.42***

0

.09

0.37***

20.11

0.15*2

0.07

9Careersatisfaction

3.61

0.78

0.09

20.15*

0

.01

20.06

20.03

0.11

0.08

0.18***

(0.86)

10Lifesatisfaction

3.56

0.82

0.07

20.07

20

.03

2

0.02

20.05

0.06

0.11

0.03

0.40***

(0.85)

11Workdev.support

3.64

0.68

0.10

0.06

20

.01

0.04

20.09

2

0.04

0.35***

0.16**

0.18**

0.02

(0.85)

12Non-workdev.

support

3.92

0.58

0.03

20.21**

0

.04

20.13*

0.00

2

0.07

0.30***

20.06

0.15*

0.17**

0.01

(0.83)

Notes:*p,

0.05;**p,

0.01;***p,

0.001.Scaleswitha-reliabilitiesalongdiagonalinparentheses;n

254

Table III.Means, standard

deviations, andcorrelations

Understandingnon-work

relationships

647

8/10/2019 Murphy & Kram 2010

12/27

Salary

Careersatisfaction

Lifesatis

faction

Variables

Model1

Model2

Model1

Model2

Model1

Model2

Gender

2

0.06

20.06

0.08

0.10

0.07

0.08

Age

0.32***

0.31***

20.17**

2

0.18**

20.10*

2

0.07*

Ethnicity

0.11*

0.11*

0.03

0.03

20.03

2

0.03

Tenureinfield

0.13*

0.15*

0.06

0.08

0.05

0.06

Industry(non-profit/Govt)

2

0.08

20.06

0.01

0.02

20.02

2

0.02

Organizationsize

0.15*

0.16**

0.12

0.12

0.05

0.07

Networksize

2

0.05

20.11

0.07

2

0.00

0.11*

0.07

Workdevelopmentalsup

port

0.14**

0.18**

0.07

Non-workdevelopmental

support

0.07

0.11*

0.15*

R2

0.24

0.31

0.06

0.12

0.05

0.08

F

10.18***

8.74***

1.91*

2.36***

2.33*

2.40**

ChangeinFa

3.09*

3.78*

2.52*

Notes:Standarizedcoefficientsreported;*p,

0.05;**p,0.0

1;***p,

0.001;aRelativetoModel1

Table IV.Salary, careersatisfaction, and lifesatisfaction regressed onwork and non-workdevelopmental networksupport

CDI15,7

648

8/10/2019 Murphy & Kram 2010

13/27

analysis (refer to Table II); this included three work developers (supervisors, uppermanagers, coworkers) and three non-work developers (parent, spouse/partner, friend).Results indicate that support from parents is positively associated with salary b0:17;p , 0:01and support from friends is positively associated with life satisfaction

b 0:16;p , 0:01.

Qualitative patterns of developmental supportAlthough the three broad developmental support functions career support,psychosocial support, and role modeling were not distinct factors quantitatively, thequalitative data indicates distinctions among functions (Kram, 1985). Specifically,participants discussed role-modeling with more frequency than career support orpsychosocial support. Four sub-functions for role-modeling were identified, whichincluded both positive and negative influences. Overall, analysis revealed differences inthe sub-functions of support emphasized by work developers versus non-work developersto facilitate individuals career success. Table V shows that non-work developersprimarily, although not exclusively, provide psychosocial support and role-modeling,whereas work developers primarily provide career support and role-modeling.

First, we address how work and non-work developers provide career and psychosocialsupport since these are well-developed functions in the mentoring literature (Kram,1985; Noe, 1988). The following section addresses the role modeling functionDifferences in career and psychosocial support between work and non-work developersParticipants most often identified career support from work relationships andpsychosocial support from non-work relationships, although both domains providesome sub-functions within each broad function. Krams (1985) career supportsub-functions of coaching and sponsorship were reported from both work andnon-work developers however, challenging assignments and exposure/visibility were

only reported from work developers and the sub-function of protection was not identified.In contrast, Krams (1985) psychosocial sub-functions of counseling and friendship werereported from both work and non-work developers. In addition, the sub-functionencouragement/emotional support, created for this study, was identified for developersfrom both domains, while the sub-function acceptance/confirmation was not identified. Aswe report these qualitative results, we also provide examples in which the role (work ornon-work) of the developer is clear in the text. For example, here is an illustration of careersupport (sub-function challenging assignments) from a work developer and psychosocialsupport (sub-function counseling) from a non-work developer:

TM was my manager and he sort of took me in and showed me the ropes and had a lot ofconfidence in me to give me projects that were probably a lot more complex than should havebeen given to somebody whos so green. That gave me the confidence to step up and toachieve beyond what I thought my capabilities were.

Outside of work I definitely always go to my wife for advice. Its easy to get frustrated atwork and sometimes it feels like youre running in place. So I guess I just go to her to vent.Shes on a different career track, but its nice to have someone to talk to.

Career support was discussed by participants in terms of coaching (43 percent),sponsorship (19 percent), challenging assignments (11 percent), and exposure/visibility(8 percent)[3]. Career support from work developers was particularly evident in

Understandingnon-work

relationships

649

8/10/2019 Murphy & Kram 2010

14/27

Broadfunction

Sub-functions

No.participants

reported

No.developers

identified

Sample

data

Careersupport

Coaching

16

17work

5non-work

Mydirectmanagerisphenomenal;hetakesthetimeto

helpmeimprove.Hesabsolutelythebestsa

lesguyIve

everworkedwithandIvelearnedmorefrom

himabout

howto

maneuveraroundsituationsandread

peopleina

strictly

businesssense,likeinmeetings,than

anyone

Sponsorship

7

7work

1non-work

Oneofthepartnersthere,oraseniormanage

ratthetime,

tookmeunderhiswingbecauseIinternedth

ere.Ilucked

outby

havinghimholdmyhandthroughm

ostofmy

internship;andactually,hefoughttogetme

aposition

afterwa

rds

Challenging

as

signments

4

7work

0non-work

He(sup

ervisor)sitsdownwithmeandaskswhatkindof

projectsIwanttodoandworkthatIwoulden

joy.Iveonly

worked

atmynewjobforsixmonthsandIvealready

donea

coupleofthingsthatwerekindofas

tretch

Exposureand

visibility

3

4work

0non-work

She(seniormanager)knewwhatpeopleInee

dedtogetin

frontofandtaughtmewhatthingsatwork

areworth

fightingforandwhicharent.Whenpositions

openedupin

othergroupsshewouldgetmynameoutinfr

ontofpeople

thatmaynotknowme,sothatwasgood

PsychosocialsupportCounselling

14

8work

8non-work

We(co-worker)definitelycomplaintoeacho

therabout

issuesatwork.Itsfrustratingdealingwithcertain

managersandwehadsimilarworksituations

aboutayear

ago.Webothjustgotnewjobs

Encouragementand

em

otionalsupport

12

3work

12non-work

Shes(s

pouse)theemotionalsupport,itsstrictly,Youcan

dothis,youregreat.Iknowyoureworking

hard.You

deserve

allthecreditthatyouget

(continued)

Table V.Examples of supportfrom work and non-workdevelopers

CDI15,7

650

8/10/2019 Murphy & Kram 2010

15/27

Broadfunction

Sub-functions

No.participants

reported

No.developers

identified

Sample

data

Friendship

4

2work

3non-work

Hesmybestfriendandhestheonetorationalizemy

thoughtprocess,justintermsofgettingawayfromwork;

letsgo

out,letsgotalk,letsgogetabeero

rsomething,

anythin

g.Dontthinkaboutworkbecauseits

amindsetin

January

thatyougotosleepseeingthenumbers

Role-modeling

Careerbehaviorto

em

ulate

24

45work

12non-work

Hewas

mydirectsupervisor,nowhesapar

tner,andhis

workingstylemadeanimpressiononme.Hepaidmuch

moreattentiontothedetailsandsupportedhisstaffin

theirdecisionswiththeclients

W

orkethicand

values

19

3work

21non-work

Mymomwenttoschoolandshefinishedcollegeinlike

two-and-a-halfyearsorsomethingcrazy.Isawherhard

workandIthinkthatsoneofthemostimpo

rtantthings

youcouldinstillinyourkids

Devaluing

re

lationships

16

17work

0non-work

Ijustseethewayhe(supervisor)managesa

ndIdont

wantto

bethatway.Theresnorecognition,

theres

favorites,theresnosenseofcaringaboutprofessional

develop

mentforhissubordinates.Iwanttokeepthese

thingsinmyheadforwhenIbecomethema

nager

W

ork-lifeinterface

fa

ilure

8

3work

7non-work

Hes(fr

iend)inthefinancialservicessidema

kinglike

$400-$5

00,000ayearandhesdonethisnow,

atthatlevel,

forprobablyfourorfiveyearsandhes36.A

ndthenon

theslip

side,nowhesgoingthroughadivorc

eandhehad

twokid

sandshelefthimbecausehewasneveraround;so

thatremindsmeaboutwork-lifebalance,itisagood

lessonforme

Notes:Quotesselectedfrommostrepresentativedomain(work

ornon-work)identifiedineachsub-fu

nction.Whendevelopersroleisunclearinthetext,

itisinsertedinparentheses

Table V.

Understandingnon-work

relationships

651

8/10/2019 Murphy & Kram 2010

16/27

relationships with supervisors (both current and former), coworkers, and upper levelmanagers. Not surprisingly, 30 percent of participants identified first supervisors asparticularly influential in their development (Kram, 1985; Berlew and Hall, 1966):

The manager of my first job, she was, at the time, Vice President of Sales. And she just kind oftook me under her wing, even when I was looking for a position out of college. I was lookingat all of my options, and she really pressed on me and said, Come here. Its a good careermove for you. And when I went there, she was a really great mentor for me.

Psychosocial support is most frequently evident in terms of counseling (38 percent),followed by encouragement and emotional support (32 percent), and friendship (11percent). Participants were more likely to discuss career and psychosocial support asintegrated or complimentary developmental components from non-work developers,especially from parents, spouse/partners, and friends, when compared to workdevelopers. This is consistent with Krams (1985) conceptualization that developmentalrelationships might provide a full range of functions and that those functions are notnecessarily mutually exclusive:

Definitely, my parents influence my career, theyre always just supportive. I can talk to thema lot about whats going on in my career and what my goals are. And then Id have to say myhusband for being so incredibly supportive of the tired and stressed out person I was while Iwent to graduate school part-time and worked full time.

We (two friends) were just always playing off each other in terms of what we think, whatcareers, and what were looking for in the next position. They have told me that they think Iwould be very good in sales and they were very encouraging. So they were probably mybiggest influence. My boyfriend was a big influence too. Hes always thought that Id be goodin sales and he suggested this path for me long ago, three years ago when we started dating.

This latter quote confirms what has recently been suggested about the value of peer

mentoring and peer coaching as mentoring alternatives (Allen and Finkelstein, 2003;Parker et al., 2008; McManus and Russell, 2007). Peer developmental relationshipsoccur in both work and non-work domains including, in this study, from coworkers,classmates, friends, and neighbors. Such relationships provide a variety ofdevelopmental functions and are characterized by mutual learning and reciprocity(Kram and Isabella, 1985).

Participants (16 percent) also discussed a lack of career support from parents due tothe changing nature of careers between generations. While their parents careers mayhave been more traditional, working for one or two organizations and movingconsistently up in a hierarchy, these participants described boundaryless careers withfrequent job changes and movement across organizations (Arthur and Rousseau, 1996)making it more difficult for their parents to understand and offer career support:

And my moms like, Its just a different world than I grew up in. So, my parents- and mydad has a great job - hes a Vice President of a small community bank - but my mom neverworked when I was growing up. Financial services, now, is just different than anything theyknew when I was growing up.

I think my father, being a big influence on me coming out of a different generation and I thinka work culture where people tended to be at jobs for a very long time. And now that Im Ithink he sees me as a bit of a job-hopper, and so thats interesting.

CDI15,7

652

8/10/2019 Murphy & Kram 2010

17/27

Differences in role-modeling between work and non-work developersQualitative data indicate that role-modeling is an equally salient form of support,comparable as a broad function to career support and psychosocial support.Participants discussed role modeling in terms of career behavior to emulate (65

percent), work ethic and values (51 percent), devaluing relationships (43 percent), andwork-life interface failure (22 percent); sub-functions which were created for this study.When discussing work relationships, participants most frequently identified rolemodels whose career behavior that they wanted to emulate. In particular, theydiscussed how their supervisor(s) or individuals they admired in upper managementinfluence how they want to behave to reach their career goals:

I think they (two upper-level managers) are driven and I definitely look at them as role-modelsif I want to succeed in this organization or any organizations. You kind of have to act howthey act, manage how they manage if you get to that point, to achieve and be a team player. Ilike the level of success that theyve achieved and how they got there over other people.

PR I looked at as a model saying, How can I be like him? rather than asking him what I need

to do to be like him. I was more of an observer, as I was just starting out and getting into moremanagerial roles. He shaped the manager Ive become.

Qualitative data also indicate that parent(s) are particularly important as role models ininfluencing participants careers. However, the 59 percent of participants who includeda parent in their developmental network discussed parental role models in terms ofvalues and work ethic rather than those whose career behavior or career path theywanted to emulate. Next, is an example of a parent role modeling values or work ethic,which participants described as shaping the way they enact their careers:

I guess my father really influences the way I look at work, especially my work ethic andadvancing. When I was a kid, probably 15 or 16, I remember him taking a day off because hewas sick, and it was the first time hed taken a day off in 25 years or something. Having that

standard of doing your job well, always showing up and doing good work, has really beeninfluenced by him.

Finally, 65 percent of participants discussed the importance of negative role models intheir developmental networks and in consciously shaping their behavior at work.Negative relationships from the work domain were most frequently a reflection ofperceived poor or unfair treatment of subordinates, clients, and customers (i.e.devaluing relationships). Participants described learning how not to behave from thesedevelopers and making sure that they did not make the same mistakes when theybecame supervisors themselves:

Shes someone from whom I learned what not to do. It was a lack of professionalism I saw inher. She was my manager, she supervised ten of us, and for instance, she would build up trust

with one employee, have them confide in her, and then turn around and actually share thatwith other people. So I learned a lot from her in terms of how to be a good leader, because Isaw what wasnt a good leader.

In contrast, negative role modeling from the non-work domain was described as afailure in navigating the work-life interface. Participants discussed friends goingthrough divorces or working too many hours to balance life outside of work.

Overall, these qualitative findings indicate that support provided by non-workdevelopers is important and that individuals attribute their own career success to these

Understandingnon-work

relationships

653

8/10/2019 Murphy & Kram 2010

18/27

relationships. These results reinforce the idea that a network of developers from bothwork and non-work domains enhances individual career success.

DiscussionTheoretical contributionsThis study has uncovered several nuances in how individuals perceive the support ofdevelopers from different domains, and empirically illustrated distinctions betweenhow work and non-work developmental relationships are related to career outcomes.While previous research suggested that developmental relationships in social arenasoutside of work were important, this study allowed for a deep empirical examination ofthe distinctive sub-functions provided by these connections. Illustrations inparticipants own words illuminated subtle, yet important, differences in the natureof support offered by individuals in the work and non-work domains. In addition,results show that support from work developers is positively associated with salarylevel and career satisfaction. In contrast, support from non-work developers is

positively associated with career satisfaction and life satisfaction.Using a mixed method approach, this study provides insight into the broadfunctions and sub-functions relevant for understanding support from developmentalnetworks that facilitates career success. Quantitative analysis indicates that thedifferent broad functions of mentoring (career support, psychosocial support, rolemodeling) have low discriminant validity when applied at the developmental networklevel. This may be due to the time intensive nature of completing a survey for an entiredevelopmental network such that participants were less inclined to discriminatebetween multiple types of support, giving each developer similar ratings across items.Alternatively, other mentoring scales applied to developmental networks may yieldmore distinguishable results, although prior studies have only yielded the two-factorsolution of career and psychosocial support (Cummings and Higgins, 2006; Higgins,

2000; Higgins and Thomas, 2001). In prior research, the function of role modeling hasbeen controversial. Empirical studies show that role modeling is a sub-function ofpsychosocial support (e.g. Ragins and Cotton, 1999), a sub-function of career support(e.g. Noe, 1988), and a broad function distinct from career and psychosocial support(e.g. Pelligrini and Scandura, 2005; Scandura, 1992). This study lends some clarity tothis debate with the inclusion of qualitative data indicating that role modeling isperceived as an equally important broad function. Specifically, role-modeling fromwork relationships is positively perceived as career behaviors to emulate andnegatively perceived as devaluing relationships (e.g. poor treatment of subordinates orclients). In contrast, role-modeling from non-work relationships is positively perceivedas shaping values and work ethic and negatively perceived as failure in navigating thework-life interface. While developmental network research has made a lot of progress

building on mentoring studies, scales designed specifically for assessing supportfunctions of developmental networks would help alleviate these issues in futureresearch.

Patterns that include non-work relationships as the most important in thedevelopmental network may reflect the instability of the twenty-first century workcontext. While supervisors and coworkers may change, it is likely that individualshave more stable and long-term relationships with people in the non-work domain (seealso, Cummings and Higgins, 2006). In particular, family member are relied upon for

CDI15,7

654

8/10/2019 Murphy & Kram 2010

19/27

unexpected functions, such as role modeling, and influence both objective andsubjective work outcomes. This may reflect the breakdown of boundaries betweenwork and family enabled by new career forms (Arthur and Rousseau, 1996; Hall, 2002;Fletcher and Bailyn, 1996). However, it may also reflect a bias in earlier mentoring

research to only include work relationships as sources of support, rather thansomething that is a new phenomenon. The qualitative data in this study contributes tothis deficiency by identifying the support functions provided by non-workrelationships most affecting individuals careers; these include the role modeling ofvalues and work ethic, counseling, encouragement and emotional support, andcoaching.

Peer mentoring relationships from both the work and non-work domains, withcoworkers and friends respectively, also play an important developmental role.Participants frequently discussed bouncing ideas off of peers, including careeropportunities and strategic decision-making. Individuals may find value in discussingideas early in their conception with peers who do not carry the same influence orexpectations as parental or supervisory relationships. It is also plausible that peerrelationships fill gaps for developmental support when individuals have negativerelationships with supervisors and provide distinct forms of psychosocial supportfrom other developers. Initiatives are already underway to foster peer coaching as adevelopmental tool, indicating that access to partnerships of mutual learning isincreasingly available (Parker et al., 2008). We anticipate that the percentage ofdevelopers that are peers is likely to increase as organizations continue to be flatter andmore team-based, and as human resource practices encourage mutual learning of thiskind. Future research is necessary to assess the full range of contributions andlimitations of peer developmental relationships.

Qualitative data show that several participants included negative workrelationships in their developmental networks. Theoretically, scholars have argued

that developmental relationships are inherently positive (Kram, 1985; Higgins andKram, 2001), thus the inclusion of negative relationships in self-reported informaldevelopmental networks was surprising. Similarly, negative mentoring relationshipshave been identified in formal programs (Eby and Allen, 2002; Ebyet al., 2004; Eby andMcManus, 2004) and negative informal role models have been examined (Gibson,2003). The inclusion of negative relationships may have depressed scores on supportscales and may be one explanation for the low explained variance in regressionequations (although these are comparable to other mentoring studies; see Allenet al.,2004). Research on developmental networks would benefit from the simultaneousinclusion of both positive and negative relationships in empirical work. It istheoretically plausible that support from developmental relationships must beconsidered in terms of optimal matching (Cutrona and Russell, 1990), with the source

and type of support matched to specific career attitudes and behaviors. It may also benecessary to analyze the developmental networks by relationship type, as post-hocanalyses indicated associations between parents support and salary as well as friendssupport and life satisfaction.

The finding that non-work developers contribute significantly to career satisfactionis important for both theory and practice. This study begins to disentangle howrelationships from different domains of developmental networks influence differentattitudes and behaviors, for example the distinction between non-work developers as

Understandingnon-work

relationships

655

8/10/2019 Murphy & Kram 2010

20/27

role models of values and work developers as role models of managerial skills. Finally,career success may not be the most appropriate outcome variable, but other constructssuch as learning or behavior towards clients and co-workers may be more relevant andinformative in todays context. Such constructs may also act as mediating or

moderating variables and add insight to the relationship between developmentalnetwork functions and career outcomes.

Practical implicationsThe broad implications of this research can assist practicing managers by increasingtheir understanding of the value of building and maintaining developmental networksthat extend beyond the boundaries of their current organization. In particular, we haveidentified the types of support individuals should seek from work AND non-work socialarenas. Thus, for example, if one is lacking in career satisfaction, it may be worthwhile toconsider multiple sources of support in both work and non-work domains. Alternatively,if one is lacking in sense of life satisfaction, then the non-work domain is where one

should enlist more support, rather than seeking it in work relationships.There are practical implications for organizations as well. Those who areresponsible for Human Resource Development processes will want to consider howwell their initiatives encourage individuals to enlist a variety of potential developers into their networks. For example, historically, one-on-one formal mentoring programshave been implemented to establish primary developmental relationships for novicesand junior executives. These data strengthen support for the value of encouragingindividuals to build multiple relationships with seniors, peers, and subordinates, aswell as with family members, friends, and other members of associations outside of thework setting (Chandler and Kram, 2010). Similarly, in the context of leadershipdevelopment programs, training in relevant relational skills as well as opportunities tobuild self-awareness will better equip individuals to take action to strengthen their own

developmental networks (Kram and Higgins, 2009; Chandler et al., 2010).Since this study reaffirms the importance of non-work relationships, individualsmay consider which relationships outside of the work domain provide encouragementand offer insight for their career development. Based on the findings that non-workdevelopers affect subjective career success, individuals should pay particular attentionto which relationships outside of work contribute to their satisfaction at work. Thisfocus may also enable individuals to capitalize on their non-work relationships assources of learning and counseling for career development. In addition, individuals inthe role of mentor might consider who in their personal lives see them as informaldevelopers and encourage discussions about careers that foster mutual learning (Kram,1985; Allenet al., 2006; Allen and Eby, 2003).

Limitations and future researchAlthough this study has several limitations, it also suggests many interestingopportunities for future research. The use of a cross-sectional survey means it isimpossible to definitively determine causality. However, the evidence that people tendto include long-term relationships in their networks and that participants ratingsindicate consistent patterns of interaction (Freeman et al., 1987) mitigates some of theinherent limitations of this research design. Due to space constraints, the surveyinstrument limited reporting of developmental network size to six relationships. Prior

CDI15,7

656

8/10/2019 Murphy & Kram 2010

21/27

research indicates that the average developmental network includes three to fiverelationships (Higgins, 2000; Higgins and Kram, 2001) and data in this study supportsthat statistic with a mean of 3.57. In addition, since the only 17 percent of participantslisted six developers, it is reasonable to assume that the data collected captures the

most important developmental relationships for participants in this sample.Retrospective interview data were collected after the survey to enrich the

interpretation of results and gain further insight into participants developmentalnetworks. The weakness of this design is the same threat to validity of all self-reportdata; that upon reflection participants may have insufficient recall as well as fabricatedor biased responses. This and other common methods bias (Doty and Glick, 1998) wasreduced with the previous collection of survey data. A longitudinal research program isnecessary to establish definitive causality between the support of developmentalrelationships and career outcomes. Career success was the outcome of interest in thisstudy; however future research might also consider how non-work developmentalrelationships contribute to other important work outcomes such as identity, workengagement, personal learning, and commitment.

The generalizability of the findings is limited by the fact that the sample was drawnfrom a population of graduate business students. As working adults attending a highlycompetitive university, all of these individuals may have achieved some degree ofcareer success already. The sample was predominantly White/Caucasian, which limitsits generalizability to other racial groups. In addition, salary level and organization sizewere assessed using categorical ranges, a conservative test because the variability ofeach measure is reduced. Future studies should compare diverse individuals in specificorganizations or industries to assess the robustness of these findings. Comparisonsacross age groups or generations and across cultures would provide more insight intothe positive and negative role of developmental networks in careers. For example, wemight expect older, more experienced individuals in later career stages to have more

diverse networks that include more peer developers in comparison with younger, lessestablished colleagues in earlier career stages.Finally, given the design of this study, it was not possible to look at how the

organizational context may have shaped individuals developmental networks. Futureresearch might examine how organizational cultures shape the extent to whichindividuals enlist developers within and outside of their organization. We expect thatsome organizations may foster networking in professional associations as well as withcustomers/clients and suppliers that could lead to developmental relationships. Otherorganizations may have impermeable boundaries that discourage outside of workdevelopers to be enlisted. Future research might compare characteristics oforganizations or industries that create various levels of opportunity forboundaryless developmental relationships and the affects of these contexts on the

diversity and range of developers as well as on critical career outcomes.

ConclusionThe opportunities for further research on developmental networks and their potentialcontributions to a wide range of career and personal outcomes are great. Similarly, theopportunities to apply this knowledge to development practices in organizations areperhaps greater than ever before due to the growing need to develop individuals atevery career stage effectively, and with limited financial resources. This study

Understandingnon-work

relationships

657

8/10/2019 Murphy & Kram 2010

22/27

highlights the distinctive contributions of both work and non-work relationships tosalary, career satisfaction, and life satisfaction. As a consequence individuals can bemore mindful of the individuals who they enlist in their developmental networks tosupport their goals, success and well-being. Similarly, organizations can create

opportunities for individuals to learn how to enlist appropriate developers in to theirnon-work networks, and to encourage access to potential developers inside and outsidethe organization through appropriate human resources practices. As organizationsstart to experiment with initiatives to foster strong developmental networks, it will beuseful for researchers to assess their impact on a wide range of both short andlong-term attitudes and behaviors so that we can continue to enhance our collectiveunderstanding of developmental networks and their potential impact. Clearly, there isstill work to be done to realize these possibilities.

Notes

1. We use the term developers to be consistent with Higgins and Krams (2001)reconceptualization of mentoring as a small group of relationships with individuals whotake an interest in the focal persons development.

2. This gender distribution reflects the participation of women in graduate managementeducation across the United States (Bureau of Labor Statistics, 2004).

3. Percentages out of total qualitative data sample,n 37.

References

Adams, G.A., King, L.A. and King, D.W. (1996), Relationships of job and family involvement,Journal of Applied Psychology, Vol. 81, pp. 411-20.

Allen, T.D. and Eby, L.T. (2003), Relationship effectiveness for mentors: factors associated withlearning and quality, Journal of Management, Vol. 29, pp. 469-86.

Allen, T.D. and Finkelstein, L.M. (2003), Beyond mentoring: alternative sources and functions ofdevelopmental support,Career Development Quarterly, Vol. 51 No. 4, pp. 346-55.

Allen, T.D., Eby, L.T. and Lentz, E. (2006), Mentorship behaviors and mentorship qualityassociated with formal mentoring programs: closing the gap between research andpractice, Journal of Applied Psychology, Vol. 91 No. 3, pp. 567-78.

Allen, T.D., Poteet, M.L. and Burroughs, S.M. (1997), The mentors perspective: a qualitativeinquiry,Journal of Vocational Behavior, Vol. 51, pp. 70-89.

Allen, T.D., Eby, L.T., Poteet, M.L., Lentz, E. and Lima, L. (2004), Career benefits associated withmentoring for proteges: a meta-analysis, Journal of Applied Psychology, Vol. 89 No. 1,pp. 127-36.

Allison, P.D. (1999), Multiple Regression: A Primer, Pine Forge Press, Thousand Oaks, CA.

Arthur, M.B. and Rousseau, D.M. (Eds.) (1996), The Boundaryless Career: A New Employment

Principle for a New Organizational Era, Oxford University Press, New York, NY.Baugh, S.G. and Fagenson-Eland, E.A. (2005), Boundaryless mentoring: an exploratory study of

the functions provided by internal versus external organizational mentors, Journal ofApplied Psychology, Vol. 35 No. 5, pp. 939-55.

Baugh, S.G. and Sullivan, E. (2005), Mentoring and career development, Career DevelopmentInternational, Vol. 10 Nos 6/7, pp. 425-8.

Berlew, D.A. and Hall, D.T. (1966), The socialization of managers: effects of expectations onperformance,Administrative Sciences Quarterly, Vol. 11 No. 2, pp. 207-23.

CDI15,7

658

8/10/2019 Murphy & Kram 2010

23/27

Cameron, K., Dutton, J. and Quinn, R. (2003),Positive Organizational Scholarship, Berrett-Koehler,San Francisco, CA.

Carlson, D.S. and Perrewe, P.L. (1999), The role of social support in the stressor-strainrelationship: an examination of work-family conflict, Journal of Management, Vol. 25

No. 4, pp. 513-40.Chandler, D. and Kram, K.E. (2010), Enlisting others in your development as a leader,

in Rothstein, M. and Burke, D. (Eds), Self-Management and Leadership Development,Edward Elgar Publishing, Northampton, MA, pp. 336-60.

Chandler, D., Hall, D.T. and Kram, K.E. (2010), A developmental network and relational savvyapproach to talent development, Organizational Dynamics, Vol. 39 No. 1, pp. 48-56.

Cohen, A. (1995), An examination of the relationships between work commitment and nonworkdomains,Human Relations, Vol. 48 No. 3, pp. 239-63.

Cohen, A. and Kirchmeyer, C. (1995), A multidimensional approach to the relation betweenorganizational commitment and nonwork participation, Journal of Vocational Behavior,Vol. 46, pp. 189-202.

Cummings, J.N. and Higgins, M.C. (2006), Relational instability at the network core: supportdynamics in developmental networks, Social Networks, Vol. 28, pp. 38-55.

Cutrona, C. and Russell, D. (1990), Type of social support and specific stress: toward a theory ofoptimal matching, in Sarason, B., Sarason, I. and Pierce, G. (Eds), Social Support:

An Interactional View, John Wiley, New York, NY, pp. 319-36.

de Janasz, S.C. and Sullivan, S.E. (2004), Multiple mentoring in academe: developing theprofessorial network, Journal of Vocational Behavior, Vol. 64 No. 2, pp. 263-83.

de Janasz, S.C., Sullivan, S.E. and Whiting, V. (2003), Mentor networks and career success:lessons for turbulent times, Academy of Management Executive, Vol. 17 No. 4, pp. 78-91.

DeLong, T.J., Gabarro, J.J. and Lees, R.J. (2008), Why mentoring matters in a hyper-competitiveworld,Harvard Business Review, Vol. 86 No. 1, pp. 115-21.

Denzin, N.K. and Lincoln, Y.S. (1994),Handbook of Qualitative Research, Sage, Thousand Oaks,CA.

Diener, E., Emmons, R.A., Larsen, R.J. and Griffin, S. (1985), The satisfaction with life scale,Journal of Personality Assessment, Vol. 49 No. 1, pp. 71-5.

Dobrow, S. and Higgins, M. (2005), Developmental networks and professional identity:a longitudinal study, Career Development International, Vol. 10 No. 6, pp. 567-83.

Doty, D.H. and Glick, W.H. (1998), Common methods bias: does common methods variancereally bias results?, Organizational Research Methods, Vol. 1 No. 4, pp. 374-406.

Dougherty, T.W. and Dreher, G.H. (2007), Mentoring and career outcomes: conceptual andmethodological issues in an emerging literature, in Ragins, B.R. and Kram, K.E. (Eds),

Handbook of Mentoring at Work: Research, Theory, and Practice, Sage, Thousand Oaks,CA, pp. 51-94.

Dutton, J.E. and Heaphy, E.D. (2003), The power of high-quality connections, in Cameron, K.S.,Dutton, J.E. and Quinn, R.E. (Eds),Positive Organizational Scholarship, Foundations of a

New Discipline, Berrett-Koehler Publishers, San Francisco, CA, pp. 263-78.

Dutton, J.E. and Ragins, B.R. (2007), Exploring Positive Relationships at Work: Building aTheoretical and Research Foundation, Lawrence Erlbaum Press, Mahwah, NJ.

Eby, L.T. (1997), Alternative forms of mentoring in changing organizational environments:a conceptual extension of the mentoring literature,Journal of Vocational Behavior, Vol. 51,pp. 125-44.

Understandingnon-work

relationships

659

8/10/2019 Murphy & Kram 2010

24/27

Eby, L.T. and Allen, T.D. (2002), Further investigation of proteges negative mentoringexperience: patterns and outcomes, Group and Organization Management, Vol. 27,pp. 456-79.

Eby, L.T. and Lockwood, A. (2005), Proteges and mentors reactions to participating in formal

mentoring programs: a qualitative investigation,Journal of Vocational Behavior, Vol. 67,pp. 441-58.

Eby, L.T. and McManus, S.E. (2004), The proteges role in negative mentoring experiences,Journal of Vocational Behavior, Vol. 65, pp. 255-75.

Eby, L.T., Butts, M., Lockwood, A. and Simon, S.A. (2004), Proteges negative mentoringexperiences: construct development and nomological validation, Personnel Psychology,Vol. 57, pp. 411-47.

Eby, L.T., McManus, S.E., Simon, S.A. and Russell, J.E.A. (2000), The proteges perspectiveregarding negative mentoring experiences: the development of a taxonomy, Journal ofVocational Behavior, Vol. 57 No. 1, pp. 1-21.

Eby, L.T., Casper, W.J., Lockwood, A., Bourdeaux, C. and Brinley, A. (2005), Work and familyresearch in IO/OB: content analysis and review of the literature (1980-2002), Journal of

Vocational Behavior, Vol. 66, pp. 124-97.Ely, R.J. (1995), The power in demography: womens social constructions of gender identity at

work,Academy of Management Journal, Vol. 38 No. 3, pp. 589-634.

Fletcher, J.K. and Bailyn, L. (1996), Challenging the last boundary: reconnecting work andfamily, in Arthur, M.B. and Rousseau, D.M. (Eds), The Boundaryless Career: A New

Employment Principle for a New Organizational Era, Oxford University Press, New York,NY.

Freeman, L.C., Romney, K. and Freeman, S.C. (1987), Cognitive structure and informantaccuracy, American Anthropologist, Vol. 89 No. 2, pp. 310-25.

Frone, M.R., Yardley, J.K. and Markel, K.S. (1997), Developing and testing an integrative modelof the work-family interface, Journal of Vocational Behavior, Vol. 50, pp. 145-67.

Gibson, D.E. (2003), Developing the professional self-concept: role model construals in early,middle, and late career stages, Organization Science, Vol. 14 No. 5, pp. 591-610.

Glaser, B. and Strauss, A. (1969), The Discovery of Grounded Theory: Strategies for QualitativeResearch, Aldine de Gruyter, New York, NY.

Gordon, J.R. and Whelan-Berry, K.W. (2005), Contributions to family and household activitiesby the husbands of midlife professional women, Journal of Family Issues, Vol. 26 No. 7,pp. 899-923.

Greenhaus, J. and Powell, G. (2006), When work and family are allies: a theory of work-familyenrichment,Academy of Management Review, Vol. 31 No. 1, pp. 72-92.

Greenhaus, J.H. and Singh, R. (2007), Mentoring and the work-family interface, in Ragins, B.R.and Kram, K.E. (Eds), Handbook of Mentoring at Work: Research, Theory, and Practice,Sage, Thousand Oaks, CA, pp. 519-44.

Greenhaus, J.H., Parasuraman, S. and Wormley, W.M. (1990), Effects of race on organizationalexperiences, job performance evaluations, and career outcomes,Academy of Management

Journal, Vol. 33 No. 1, pp. 64-86.

Hall, D.T. (2002), Careers in and out of Organizations, Sage, Thousand Oaks, CA.

Heslin, P.A. (2005), Conceptualizing and evaluating career success, Journal of OrganizationalBehavior, Vol. 26, pp. 113-36.

Higgins, M.C. (2000), The more, the merrier? Multiple developmental relationships and worksatisfaction, Journal of Management Development, Vol. 19 No. 4, pp. 277-96.

CDI15,7

660

8/10/2019 Murphy & Kram 2010

25/27

Higgins, M.C. (2001), Changing careers: the effects of social context, Journal of OrganizationalBehavior, Vol. 22, pp. 595-618.

Higgins, M.C. (2005), Developmental Network Questionnaire, Harvard Business School Press,Cambridge, MA.

Higgins, M.C. and Kram, K.E. (2001), Reconceptualizing mentoring at work: a developmentalnetwork perspective, Academy of Management Review, Vol. 26 No. 2, pp. 264-88.

Higgins, M.C. and Thomas, D.A. (2001), Constellations and careers: toward understanding theeffects of multiple developmental relationships, Journal of Organizational Behavior,Vol. 22, pp. 223-47.

Holsti, O. (1969), Content Analysis for the Social Sciences and Humanities , Addison-Wesley,Reading, MA.

Ibarra, H. (1992), Homophily and differential returns: sex differences in network structure andaccess in an advertising firm,Administrative Science Quarterly, Vol. 37 No. 3, pp. 422-47.

Ibarra, H. (1995), Race, opportunity, and diversity of social circles in managerial networks,Academy of Management Journal, Vol. 38 No. 3, pp. 673-704.

Judge, T.A., Bono, J.E., Erez, A. and Locke, E.A. (2005), Core self-evaluations and job and lifesatisfaction: the role of self-concordance and goal attainment, Journal of Applied

Psychology, Vol. 90 No. 2, pp. 257-68.

Judge, T.A., Cable, D.M., Boudreau, J. and Wand Bretz, R.D. (1995), An empirical investigationof the predictors of executive career success, Personnel Psychology, Vol. 48, pp. 485-519.

Kirchmeyer, C. (1992), Nonwork participation and work attitudes: a test of scarcity vs expansionmodels of personal resources, Human Relations, Vol. 45, pp. 775-95.

Kirchmeyer, C. (1993), Nonwork-to-work spillover: a more balanced view of the experiences andcoping of professional women and men,Sex Roles, Vol. 28, pp. 531-52.

Kram, K.E. (1985),Mentoring at Work: Developmental Relationships in Organizational Life, Scott,Foresman & Company, Glenville, IL.

Kram, K.E. (1996), A relational approach to career development, in Hall, D.T. (Ed.), The CareerIs Dead Long Live the Career: A Relational Approach to Careers, Jossey-Bass, SanFrancisco, CA, pp. 132-57.

Kram, K.E. and Higgins, M.A. (2009), A new mindset on mentoring: creating developmentalnetworks at work,Sloan Management Review, Cambridge, MA.

Kram, K.E. and Isabella, L.A. (1985), Mentoring alternatives: the role of peer relationships incareer development, Academy of Management Journal, Vol. 28, pp. 110-32.

Landau, J. and Arthur, M.B. (1992), The relationship of marital status, spouses career status,and gender to salary level, Sex Roles, Vol. 27, pp. 665-81.

Lee, T.W., Mitchell, T.R., Sablynski, C.J., Burton, J.P. and Holtom, B.C. (2004), The effects of job

embeddedness on organizational citizenship, job performance, volitional absences, andvoluntary turnover, Academy of Management Journal, Vol. 47 No. 5, pp. 711-22.

Lent, R.W. and Brown, S.D. (1996), Social cognitive approach to career development:an overview, Career Development International, Vol. 44 No. 4, pp. 310-22.

Lent, R.W., Singley, D., Sheu, H., Gainor, K.A., Brenner, B.R., Treistman, D. and Ades, L. (2005),Social cognitive predictors of domain and life satisfaction: exploring theoretical precursorsof subjective well-being,Journal of Cognitive Psychology, Vol. 52 No. 3, pp. 429-42.

Levinson, D. (1978), The Seasons of a Mans Life, Alfred A. Knopf, New York, NY.

Understandingnon-work

relationships

661

8/10/2019 Murphy & Kram 2010

26/27

McManus, S. and Russell, J. (2007), Peer mentoring relationships, in Ragins, B.R. andKram, K.E. (Eds),Handbook of Mentoring at Work: Research, Theory, and Practice , Sage,Thousand Oaks, CA, pp. 273-98.

Miles, M.B. and Huberman, A.M. (1994), Qualitative Data Analysis: An Expanded Sourcebook,

2nd ed., Sage, Thousand Oaks, CA.

Molloy, J.C. (2005), Developmental networks: literature review and future research, CareerDevelopment International, Vol. 10 No. 6, pp. 536-47.

Noe, R.A. (1988), An investigation of the determinants of successful assigned mentoringrelationships,Personnel Psychology, Vol. 41, pp. 457-79.

Parasuraman, S., Greenhaus, J.H. and Granrose, C.S. (1992), Role stressors, social support, andwell-being among two-career couples, Journal of Organizational Behavior, Vol. 13 No. 4,pp. 339-56.

Parker, P., Hall, D.T. and Kram, K.E. (2008), Peer coaching: a relational process for acceleratingcareer learning, Academy of Management Learning and Education, Vol. 7 No. 4,pp. 487-503.

Pellegrini, E.K. and Scandura, T.A. (2005), Construct equivalence across groups: an unexploredissue in mentoring research, Educational and Psychological Measurement, Vol. 65 No. 2,pp. 323-35.

Raabe, B. and Beehr, T.A. (2003), Formal mentoring versus supervisor and co-workerrelationships: differences in perceptions and impact,Journal of Organizational Behavior,Vol. 24 No. 3, pp. 271-93.

Ragins, B.R. and Cotton, J.L. (1999), Mentor functions and outcomes: a comparison of men andwomen in formal and informal mentoring relationships, Journal of Applied Psychology,Vol. 84, pp. 529-50.

Ragins, B.R. and Kram, K.E. (2007), Handbook of Mentoring at Work: Research, Theory, andPractice, Sage, Thousand Oaks, CA.

Reay, T., Golden-Biddle, K. and Germann, K. (2006), Legitimizing a new role: small wins andmicroprocesses of change, Academy of Management Journal, Vol. 49 No. 5, pp. 977-98.

Rothbard, N.P. (2001), Enriching or depleting? The dynamics of engagement in work and familyroles,Administrative Science Quarterly, Vol. 46, pp. 655-84.

Sandelowski, M. (1995), Sample size in qualitative research, Research in Nursing and Health,Vol. 18 No. 2, pp. 179-83.

Scandura, T.A. (1992), Mentorship and career mobility: an empirical investigation, Journal ofVocational Behavior, Vol. 43, pp. 251-65.

Seibert, S.E., Kraimer, M.L. and Liden, R.C. (2001), A social capital theory of career success,Academy of Management Review, Vol. 44 No. 2, pp. 219-37.

Sluss, D.M. and Ashforth, B.E. (2007), Relational identity and identification: defining ourselvesthrough work relationships, Academy of Management Review, Vol. 32 No. 1, pp. 9-32.

Strauss, A. and Corbin, J. (1990), Basics of Qualitative Research: Grounded Theory Proceduresand Techniques, Sage, London.

van Emmerik, I.J.H. (2004), The more you can get, the better: mentoring constellations andintrinsic career success, Career Development International, Vol. 9 No. 6, pp. 578-94.

Wimpenny, P. and Gass, J. (2000), Interviewing in phenomenology and grounded theory: is therea difference?, Journal of Advanced Nursing, Vol. 31 No. 6, pp. 1485-92.

CDI15,7

662

8/10/2019 Murphy & Kram 2010

27/27

About the authorsWendy Marcinkus Murphy is an Assistant Professor of Management in the College of Businessat Northern Illinois University. She received a PhD in organizational studies from BostonCollege. Murphy has research interests in the areas of developmental relationships, careers,

identity, and the work-life interface. She has published her work in Women in ManagementReview and presented her research at both national and regional professional conferences.Wendy Marcinkus Murphy is the corresponding author and can be contacted at: [email protected]

Kathy E. Kram is the Shipley Professor in Management in the Department of OrganizationalBehavior at Boston University. She received her PhD from Yale University. She has researchinterests in the areas of adult development, mentoring, diversity in executive development,leadership, and organizational change processes. In addition to her book,Mentoring at Work, shehas published in a wide range of journals. Her latest book, an edited volume with Professor BelleRose Ragins, is entitled Handbook of Mentoring at Work: Theory, Research and Practice .

Understandingnon-work

relationships

663

To purchase reprints of this article please e-mail: [email protected] visit our web site for further details: www.emeraldinsight.com/reprints