Top Banner
Decision Sciences Volume 35 Number 3 Summer 2004 Printed in the U.S.A. Evaluating the Deming Management Model of Total Quality in Services Thomas J. Douglas and Lawrence D. Fredendall Department of Management, Clemson University, 101 Sirrine Hall, Clemson, SC 29634-1305, e-mail: [email protected], fl[email protected] ABSTRACT This article uses the Deming management model developed by Anderson et al. (1994b) as an initial template to analyze total quality in services. While the literatures addressing quality management have developed separately for products and services, the founders of total quality portrayed this management philosophy as universally oriented. Our study first replicates two earlier studies that tested the Deming management model in manu- facturing industries. Using hospitals as our unit of analysis, we realized findings similar to the earlier manufacturing studies. Next, we used contributions from the MBNQA literature to test an enhanced model. Our subsequent findings support the MBNQA con- cept that “leadership drives the system that creates results” and provides evidence of the ubiquitous importance of leadership for ensuring the success of a quality improvement program. Finally, an anomaly of this study and those published earlier is the inabil- ity to find support for the relationship between continuous improvement and customer satisfaction. Integrating the substantial work in the service quality literature focused on customer satisfaction measurement is recommended to future researchers to help resolve this issue and further enhance the model. Subject Areas: Deming Model, Service Quality, and Total Quality Management. INTRODUCTION Services are the dominant segment in the economy of the United States. The per- centage of workers involved in the service industries in the United States has increased from 30 percent in 1900 to over 85 percent in 2000 (Bureau of Labor Statistics, 2002). In addition, many manufacturing companies provide services as well as products (e.g., automobiles and service dealerships). Despite services being a large segment of the economy, the concepts of ser- vice quality are not as well developed as those of manufacturing quality (Ghoba- dian, Speller, & Jones, 1994). This may be because the manufacturing and service We thank the special issue guest editor, Kurt Bretthauer, and two anonymous reviewers for their very constructive feedback. We are also grateful to Joel Ryman for his contribution of key performance data and to Bill Judge for his insights and thoughtful comments in the design of the original study. An earlier version of this article was presented at the 2003 Academy of Management Meetings in Seattle. Corresponding author. 393
31
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Deming Model

Decision SciencesVolume 35 Number 3Summer 2004Printed in the U.S.A.

Evaluating the Deming Management Modelof Total Quality in Services∗

Thomas J. Douglas and Lawrence D. Fredendall†Department of Management, Clemson University, 101 Sirrine Hall, Clemson, SC 29634-1305,e-mail: [email protected], [email protected]

ABSTRACT

This article uses the Deming management model developed by Anderson et al. (1994b)as an initial template to analyze total quality in services. While the literatures addressingquality management have developed separately for products and services, the foundersof total quality portrayed this management philosophy as universally oriented. Our studyfirst replicates two earlier studies that tested the Deming management model in manu-facturing industries. Using hospitals as our unit of analysis, we realized findings similarto the earlier manufacturing studies. Next, we used contributions from the MBNQAliterature to test an enhanced model. Our subsequent findings support the MBNQA con-cept that “leadership drives the system that creates results” and provides evidence of theubiquitous importance of leadership for ensuring the success of a quality improvementprogram. Finally, an anomaly of this study and those published earlier is the inabil-ity to find support for the relationship between continuous improvement and customersatisfaction. Integrating the substantial work in the service quality literature focused oncustomer satisfaction measurement is recommended to future researchers to help resolvethis issue and further enhance the model.

Subject Areas: Deming Model, Service Quality, and Total QualityManagement.

INTRODUCTION

Services are the dominant segment in the economy of the United States. The per-centage of workers involved in the service industries in the United States hasincreased from 30 percent in 1900 to over 85 percent in 2000 (Bureau of LaborStatistics, 2002). In addition, many manufacturing companies provide services aswell as products (e.g., automobiles and service dealerships).

Despite services being a large segment of the economy, the concepts of ser-vice quality are not as well developed as those of manufacturing quality (Ghoba-dian, Speller, & Jones, 1994). This may be because the manufacturing and service

∗We thank the special issue guest editor, Kurt Bretthauer, and two anonymous reviewers for their veryconstructive feedback. We are also grateful to Joel Ryman for his contribution of key performance data andto Bill Judge for his insights and thoughtful comments in the design of the original study. An earlier versionof this article was presented at the 2003 Academy of Management Meetings in Seattle.

†Corresponding author.

393

Page 2: Deming Model

394 Evaluating the Deming Management Model

literatures currently treat quality management differently (Harvey, 1998; Sousa &Voss, 2002). Most service quality research uses the gap model to examine servicequality (Harvey, 1998). The gap model, which was first proposed by Parasuraman,Zeithaml, and Berry (1985), considers five gaps between service performance andcustomer expectations, but it does not directly consider many of the elements oftotal quality.

The views of the founders of total quality management (i.e., Deming, Juran,and Ishikawa) are prominent in the manufacturing literature, but not in services.However, the theoretical foundation and methods of total quality support its usefor both products and services (Anderson, Rungtusanatham, & Schroeder, 1994b;Waldman, 1994). In fact, Deming (1986) devoted an entire chapter of his book,Out of Crisis, to service industries. In addition, the Malcolm Baldrige NationalQuality Awards (MBNQA) program, established by the U.S. Congress in 1987,encompassed seven categories that could be applied to any organization, whetherin manufacturing or services (Bell & Keys, 1998).

This article is a first step toward integrating the service quality and productquality literatures. The article uses a conceptually based, empirically tested modelto evaluate potential commonalities between quality management concepts in man-ufacturing and service environments. First, the study demonstrates that the totalquality or Deming management model developed by Anderson et al. (1994b), andtested in manufacturing industries (see Anderson, Rungtusanatham, Schroeder, &Devaraj, 1995; Rungtusanatham, Forza, Filippini, & Anderson, 1998), is theoret-ically applicable to services. Second, this article tests the Deming managementmodel with data from hospitals. Third, the study incorporates contributions fromtwo additional bodies of literature. It uses the MBNQA and the service qualityliteratures to suggest further enhancements for the Deming management model.Fourth, the implications of the empirical findings are examined for both servicesand manufacturing.

PRESENTATION AND EVALUATIONOF THE DEMING MANAGEMENT MODEL

The definition of each construct underlying the Deming management model, as for-mulated by Anderson et al. (1994b), is displayed in Table 1. These seven constructswere developed from extensive readings of published materials by W. EdwardsDeming as well as other quality experts. Insights from the readings were furtherdeveloped via a three-round Delphi study using academic and practitioner expertsto identify the concepts underlying Deming’s 14 points (Deming, 1986). The re-search team then clustered these concepts into seven constructs to represent thecontent of the Deming management method.

To establish the credibility of the seven constructs in the Deming manage-ment method, each construct was compared to the existing management literature(Anderson et al., 1994b). All of the constructs were supported by this literature.The relationships between the constructs were developed into theory, displayed inFigure 1, using a relations diagram (e.g., Goal/QPC, 1991).

There have been two empirical examinations of the Deming managementmodel in the literature. First, Anderson et al. (1995), using measures identified from

Page 3: Deming Model

Douglas and Fredendall 395

Table 1: Constructs underlying the Deming management method (Anderson et al.,1994b, p. 480).

Visionary Leadership The ability of management to establish, practice, and lead along-term vision for the organization, driven by changingcustomer requirements, as opposed to an internalmanagement control role.

Internal and ExternalCooperation

The propensity of the organization to engage in noncompetitiveactivities internally among employees and externally withrespect to suppliers.

Learning The organizational capability to recognize and nurture thedevelopment of its skills, abilities, and knowledge base.

Process Management The set of methodological and behavioral practicesemphasizing the management of process, or means ofactions, rather than results.

ContinuousImprovement

The propensity of the organization to pursue incremental andinnovative improvements of its processes, products, andservices.

Employee Fulfillment The degree to which employees of an organization feel that theorganization continually satisfies their need.

Customer Satisfaction The degree to which an organization’s customers continuallyperceive that their needs are being met by the organization’sproducts and services.

Figure 1: Proposed theory of quality management for services.

Organizational

System

Process

Outcomes

Internal and

External

Cooperation

Continuous

Improvement

Visionary

Leadership

Process

Management

Customer

Satisfaction

Learning Employee

Fulfillment

the world-class manufacturing research project (Flynn, Schroeder, & Sakakibara,1994), found exploratory support for most of the hypothesized relationships.Rungtusanatham et al. (1998) replicated the first study using manufacturing fa-cilities located in Italy. Both the Anderson et al. (1995) and Rungtusanatham et al.(1998) studies supported most of the relationships in the Deming managementmodel. However, both studies suggested that additional testing of the Demingmanagement model in other industries was necessary.

Page 4: Deming Model

396 Evaluating the Deming Management Model

To evaluate and test this model in hospitals, we will first evaluate the con-structs and relationships of the model to determine their relevance to service indus-tries in general and to hospitals in the health care industry specifically. In addition,we will investigate the theoretical and empirical contributions suggested by therecent studies depicting and evaluating the theoretical framework underlying theMBNQA criteria and their relationship to the Deming management model. Thisis appropriate since the MBNQA criteria were designed to comprehensively coveroverall management requirements from the perspective of the diverse quality com-munity (Bell & Keys, 1998).

Service and Health Care Literatures

Each construct in the Deming management model, as depicted in Figure 1, isexamined separately in the following subsections to determine its applicabilityto the service and health care industries. This is accomplished by reviewing theexisting service quality and health care literatures and the MBNQA criteria.

Visionary leadership

There is mixed support in the service quality literature for the visionary leadershipconstruct. While Harvey’s (1998) review of service quality does not consider therole of leadership, Chase (1996) reports evidence that superior leadership leadsto superior results in services. In addition, Foster, Howard, and Shannon (2002)found that leadership was related to process improvement, teamwork, and employeesatisfaction in their analysis of government services.

Several health care articles addressed the importance of visionary leader-ship in quality program implementation (e.g., Arndt & Bigelow, 1995; Bender& Krasnick, 1995; Huq & Martin, 2000; Motwani, Sower, & Brashier, 1996).Motwani et al. (1996) reviewed the quality implementation literature for hospi-tals and suggested an implementation model led by top management. Arndt andBigelow (1995, p. 7) stated that top management had “the responsibility of imbu-ing organizational members with a shared sense of purpose aimed at creating bothvalue for the customer and the committed involvement of organizational employ-ees.” However, they also argued that top administrators in hospitals might not haveresponsibility for some of the key organizational activities, such as those controlledby physicians.

Finally, the importance of leadership in successfully implementing total qual-ity in health care is recognized in the MBNQA’s health care criteria for performanceexcellence (NIST, 2003). The first award criterion listed is “visionary leadership”and it outlines the key role of senior administrators in successfully implementingtotal quality in health care. Therefore, while the concept of visionary leadershipmay be difficult to implement in hospitals due to their organizational structure, itremains an important element of total quality management.

Internal and external cooperation

Both the service quality literature and the health care literature support the appli-cability of the internal and external cooperation construct to the service industry.The internal cooperation construct is similar to the concept of human resource

Page 5: Deming Model

Douglas and Fredendall 397

focus as used in service research. Typical items included in human resource fo-cus are communications, training, recognition of support for quality objectives,and employee satisfaction (e.g., Cook & Verma, 2002). These are very similar tothe measures used by Anderson et al. (1995), which included communication andemployee involvement in decisions.

The external cooperation concept should be valid for those services that usesignificant supplier inputs. For example, Callahan and Moretton (2001) found thatexternal involvement of sales and marketing with their suppliers shortened soft-ware development time. Gittell (2002) examined provider-provider relationshipsin health care, that is, relationships between service providers who jointly pro-vide service to the patient, and found that increased internal coordination led toincreased customer satisfaction. For example, Douglas and Ryman (2003) arguedthat relationships between hospitals and independent physician groups enabled thecontinuum of care provided to patients to be more efficient and effective given theutilization of complementary resources. Gittell (2002) also concluded that serviceoperations, which involve reciprocal interdependence, require a high coordinationlevel between service providers. Carmen et al. (1996) provided an in-depth re-view and analysis of 10 hospitals and concluded that a culture supportive of groupaffiliation and teamwork led to better quality results, including customer satisfac-tion. Finally, the MBNQA’s health care criteria, in the section on valuing staffand partners, discuss the importance of building partnerships, both internally andexternally, in a quality program.

Learning

Since most services are delivered through the contact of service personnel with thecustomer, employee learning should have a significant impact on service quality.However, there has been little empirical research about the impact of learning onservice quality. Roth and Jackson (1995) measured the operations capabilities ofservice organizations as their organizational knowledge (i.e., institutional knowl-edge expressed as system design choices and competent knowledgeable staff),technological leadership, market acuity, and factor productivity. They found thatall of these operations capabilities and market acuity influenced service quality.

Huq and Martin (2000) found that the education and training of the healthcare workforce in basic quality principles, statistics, and interpersonal skills, andthe creation of a learning environment were important for successful quality imple-mentations in hospitals. In addition, Bender and Krasnick (1995) argued that whileorganizational learning was a critical aspect of quality management in hospitals,the typical health care organization did not provide a culture that allowed currentconcepts and assumptions to be challenged. Therefore, even though the ability tolearn was important for quality management in hospitals, many hospitals’ cultureslessened the ability to learn.

The MBNQA’s health care criteria include organizational and personal learn-ing. These concepts are defined broadly, encompassing continuous learning withinthe organization and the ability to adapt to environmental change. The expected out-comes of successfully embracing this concept include efficiency, responsiveness,and better performance.

Page 6: Deming Model

398 Evaluating the Deming Management Model

Process management

Anderson et al. (1994b) viewed process management as being those practices thatfocus on managing the means or actions taken in the process and not the results.The applicability of the process management construct in the Deming managementmodel to services is supported by both the service and hospital literatures. There arethree major bodies of service process research. The first considers how to design aservice process, using such tools as service blueprinting (e.g., Shostack, 1984). Thesecond considers how customer contact is managed and how it influences the serviceprocess and the customers’ perception of quality (e.g., Kellogg & Chase, 1995).The third examines quality management techniques and procedures such as the useof statistical quality control (e.g., Sureshchandar, Rajendran, & Anantharaman,2001) and mistake proofing (e.g., Stewart & Chase, 1999) in services.

The hospital literature also recognizes the importance of process managementto service quality. For example, Shortell et al. (1995) included a hospital’s focus onmanaging processes and systems in measuring its continuous quality improvement(CQI)/total quality management (TQM) program, the implementation of which wasrelated to better perceived patient outcomes. In a study of 227 European hospitals,Kunst and Lemmink (2000) found that process management, as defined in theEuropean Quality Awards, was positively related to quality program outcomesrepresenting both customer and employee satisfaction. The importance of processmanagement to service quality is recognized by the MBNQA program in healthcare, which includes process management as one of its seven criteria.

Continuous improvement

There is limited support for the continuous improvement construct in the gen-eral service quality literature, but strong support in the health care literature.Sureshchandar et al. (2001) state that striving for continuous improvement is criticalto the achievement of service quality. In addition, Roth and Jackson (1995) foundan important role for continuous improvement in the firm’s ability to deliver highservice quality. They operationalized continuous improvement as organizationalknowledge, and found that organizational knowledge was the primary determinantof service quality.

Hospitals often have Continuous Quality Improvement (CQI) programs.Carmen et al. (1996) argue that the key for hospital success is a commitmentto continuous improvement. Routhieaux and Gutek (1998) measure the effective-ness of CQI using a framework that includes continuous improvement based ona customer focus. The MBNQA’s criteria measure the support of the continuousimprovement of processes and, ultimately, systems.

Hospitals receive external pressure for continuous improvement from theJoint Commission on Accreditation of Health Care Organizations (JCAHO). TheJCAHO conducts periodic reviews, which focus on documented evidence of con-tinuous improvement efforts (Westphal, Gulati, & Shortell, 1997).

Employee fulfillment

Anderson et al. (1994b) suggest that employee fulfillment be measured by job satis-faction, job commitment, and pride in their work. The relevance of this construct to

Page 7: Deming Model

Douglas and Fredendall 399

service quality is strongly supported. Sureshchandar et al. (2001) stress that firmsmust focus on employee satisfaction because there is a high correlation betweenemployee perceptions of well-being and the customers’ perception of service qual-ity. Hensel (1990) also found that service employee participation leads to employeeownership of the service being delivered and that the quality challenge for servicecompanies is to create the correct environment for the employees to function within.

The hospital service literature suggests that employee fulfillment is necessaryfor enhanced performance outcomes, including customer satisfaction, but may bedifficult to achieve. Arndt and Bigelow (1995) argued that gaining the committedinvolvement of employees may be difficult in hospitals. However, Boerstler et al.(1996) found that hospitals that ranked high on employee participation, teamwork,and an adaptable, flexible culture were most successful in the quality managementinitiatives. Huq and Martin (2000) echo these findings and add the need to have em-powered employees with authority and responsibility to ensure successful qualityimplementation. Shortell et al. (1995) found that successful quality implementa-tions led to increased employee relations and satisfaction.

Customer satisfaction

Customer satisfaction is a very important concept in the service quality literature,addressing how well the service and the service process meet customer expec-tations (Harvey, 1998). Nilsson, Johnson, and Gustafsson (2001) found that theinternal quality practices of service firms influence customer satisfaction. In ad-dition, Anderson, Fornell, and Lehmann (1994a) found that, in a service firm,customer satisfaction influences a firm’s profitability, while Rust, Zahorik, andKeiningham (1995) found that increased customer satisfaction in services leads tohigher customer retention.

Organizational processes in health care focus on the total satisfaction ofcustomers, that is, patients (Zabada, Rivers, & Munchens, 1998). Most articlesaddressing quality management in health care list multiple outcomes, with patientsatisfaction as a key outcome. For instance, Arndt and Bigelow (1995) measureimprovement in efficiency and effectiveness in terms of cost controls, clinicaloutcomes, and satisfaction for all customers including patients, their families, andall internal constituents. Shortell et al. (1995) and the MBNQA program criteriameasure very similar categories.

Hypotheses

The review of the literature established the validity of the constructs in Figure 1for services in general and health care in particular. In this section, we examine therelationships between the constructs and develop formal hypotheses for testing.

Visionary leadership is core to the Deming management model. Leadershipis essential in order to create a service organization that has both internal andexternal cooperation. Deming (1986) stressed the need to eliminate fear to improvethe process. Leadership is the primary determinant of the level of fear, whichcontrols the level of cooperation. This was supported by Anderson et al. (1995)and Rungtusanatham et al. (1998) for manufacturing plants and the same rationaleseems to apply to services. So, it is hypothesized that:

Page 8: Deming Model

400 Evaluating the Deming Management Model

Hypothesis 1: Visionary leadership is positively related to internal andexternal cooperation.

Leaders control organizational learning by allocating the resources and choos-ing whether or not to recognize and reward the learning that occurs. This link wassupported by both Anderson et al. (1995) and Rungtusanatham et al. (1998) formanufacturing and should be true for services as well. So, it is hypothesized that:

Hypothesis 2: Visionary leadership is positively related to learning.

Internal and external cooperation was found to be important to process man-agement in manufacturing by Anderson et al. (1995) and Rungtusanatham et al.(1998). Internal and external cooperation should also be essential to service processmanagement. Internal cooperation should facilitate data sharing, the standardiza-tion of processes, the visual tracking of defects, and the use of statistical toolsto identify problems, all emphasizing the management of the process. So, it ishypothesized that:

Hypothesis 3: Internal and external cooperation is positively relatedto process management.

The model in Figure 1 also proposes that learning leads to process man-agement. This link was not supported by either the Anderson et al. (1995) or theRungtusanatham et al. (1998) study. However, the scales in their studies seemedto measure only the amount of task training employees received. This study mea-sures the total quality training the employees receive and how the organization usescustomer information. It is believed that total quality training allows employeesto learn from and respond to available customer information. This learning fromcustomer information should lead to process management, which includes makingdata available and using data to evaluate performance. So, it is hypothesized that:

Hypothesis 4: Learning is positively related to process management.

The path from process management to continuous improvement was sup-ported by both the Anderson et al. (1995) and Rungtusanatham et al. (1998) stud-ies of manufacturing. It should also be significant in services. There is pressure onservices to continuously improve to meet rising customer expectations. It is arguedby many quality management researchers that process management is essentialto improvement. In this context, process management means analyzing currentperformance and taking action. Without taking actions based on the analysis, im-provement is unlikely. So, it is hypothesized that:

Hypothesis 5: Process management is positively related to continuousimprovement.

The path from process management to employee fulfillment was only weaklysupported by Anderson et al. (1995) and not supported by Rungtusanatham et al.(1998) in manufacturing. Their scales measured employee pride in their work.We view employee fulfillment in terms of employees being given the authority,information, and tools necessary to do their job. Since service personnel mustuse the processes available to them to respond quickly to and satisfy individual

Page 9: Deming Model

Douglas and Fredendall 401

customer needs, employees’ fulfillment depends on superior process management.So, it is hypothesized that:

Hypothesis 6: Process management is positively related to employeefulfillment.

Figure 1 proposes that customer satisfaction is the result of both continu-ous improvement and employee fulfillment. The link from continuous improve-ment to customer satisfaction was not supported by Anderson et al. (1995) butwas marginally supported by Rungtusanatham et al. (1998). It is expected thatcontinuous improvement in services should lead to customer satisfaction. In ourstudy, business performance is measured in three ways—customer satisfaction,perceived financial performance, and the JCAHO audit score. Continuous improve-ment should lead to both improved financial performance and improved JCAHOaudit scores. Continuous improvement will do this by reducing waste, thus cuttingcosts and improving finances. Continuous improvement should also improve auditscores, since audits are so important to hospitals and top management is under in-stitutional pressure to do well on the audits (Westphal et al., 1997). The MBNQAcriteria also include similar, multiple measures of business performance (Meyer &Collier, 2001). So, it is hypothesized that:

Hypothesis 7: Continuous improvement is positively related to busi-ness performance.

The link from employee fulfillment to customer satisfaction was supportedby Anderson et al. (1995) and not supported by Rungtusanatham et al. (1998)in manufacturing. This link should be supported for services. Organizations thatallow employees to adapt to the customer’s needs using available processes andinformation should perform better. It is believed that employee fulfillment will helpimprove financial performance by leading to improved service, which enhances thehospital’s reputation. This reputation should in turn lead to increased business atthe hospital. Employee fulfillment can improve audit scores, as employees take theinitiative to eliminate obstacles, which might prevent fulfilling the requirements ofthe JCAHO audits. So, it is hypothesized that:

Hypothesis 8: Employee fulfillment is positively related to businessperformance.

Measuring Framework Constructs and Other Variables

The constructs underlying the Deming management model were operationalizedusing previously published scales in the TQM literature. When necessary, multiplescales were used (e.g., internal and external cooperation). Table 2 lists the sourceof the scales used to represent each construct, and displays their reliability scoresbased on our study data.

The reliability and validity of these scales was extensively tested by theircreators. Additional factor and reliability analyses were conducted in this study,since the number of items in most scales was reduced for efficiency reasons and thisis a new data set. The scales and associated items are contained in the Appendix.

Page 10: Deming Model

402 Evaluating the Deming Management Model

Table 2: Theoretical constructs, measurement scales, sources, and reliabilities.

ReliabilityConstruct Scale Source (Alpha)

Visionary Leadership Top Management TeamInvolvement (6 items)

Saraph et al. (1989) .92

Internal and ExternalCooperation

Quality Philosophy (5 items) Zeitz et al. (1997) .85Supplier Involvement

(6 items)Saraph et al. (1989) .89

Learning Total Quality Training(6 items)

Saraph et al. (1989) .91

Customer Driven (5 items) Zeitz et al. (1997) .87Powell (1995)

Process Management Management by Fact (6 items) Saraph et al. (1989) .92Total Quality Methods

(5 items)Zeitz et al. (1997) .87

ContinuousImprovement

Continuous Improvement(3 items)

Zeitz et al. (1997) .90

Employee Fulfillment Structural exploration(2 items)

Khandwalla (1977) .67

As displayed in Table 2, the scale used in this study to measure visionaryleadership was initially developed by Saraph, Benson, and Schroeder (1989) tomeasure top management team involvement. It was modified slightly to ensureits suitability for hospitals, but is very similar to the scales used by Andersonet al. (1995) in their study of the Deming management model. The scale fo-cuses on the roles and participation of top management in leading the qualityinitiative.

The scales to measure internal and external cooperation were different thanthose used by Anderson et al. (1995). Two scales were used. The first scale, qualityphilosophy, is used as a measure of internal cooperation. It measures commitmentand awareness of the firm’s mission and management’s role in preventing prob-lems. Supplier involvement measures such things as knowledge of the supplierand whether the hospital offers a long-term relationship to the suppliers. All of theitems assume that we have a cooperative relationship with the supplier. Use of bothscales allows us to capture the compete domain of this construct.

Anderson et al. (1995) measured learning as the amount of training the em-ployees received. This article uses two different scales to measure learning. Thefirst looks at the total quality training given to employees. This seems to be amore appropriate measure in the context of the Deming management model interms of providing the skills needed to implement quality. The second scale,titled “Customer Driven Information,” measures what employees do to learnabout their customers’ needs, increasing their knowledge and ability to meetthese needs. Combined, these scales map well into the learning construct in thismodel.

The process management construct was measured with two scales. The firstscale, management by fact, has some similarity to the scale used by Anderson et al.

Page 11: Deming Model

Douglas and Fredendall 403

(1995), but it also measures who receives and uses quality data and whether thedata are used to manage quality. The second scale, total quality methods, measureswhether the employees use the quality tools to manage their process. This is a scaledeveloped to measure total quality practices (Zeitz, Johannesson, & Ritchie, 1997)actually used in an organization, and is an appropriate measure of the extent towhich these are used in the daily process. Both scales focus on the availability ofinformation and the actions taken to manage processes.

The continuous improvement construct was measured using a scale that wasvery similar to that used by Anderson et al. (1995). The employee fulfillment con-struct was measured by Anderson et al. (1995) as the amount of pride the employeetakes in their work. A different approach was taken in this study. The structuralexploration scale used here to measure employee fulfillment was developed byKhandwalla (1977) as a measure of organization design. It was felt that employeeswho had access to the necessary data and were allowed to adapt as necessary tocomplete their task requirements would be fulfilled. This is supported by a model ofempowered behavior that argues that resource and information sharing in the localwork environment leads to psychological empowerment and empowered behavior(Robbins, Crino, & Fredendall, 2002).

While the Deming management model contains three outcome variables(continuous improvement, employee fulfillment, and customer satisfaction), weadded two more variables pertinent to hospitals, financial performance and overallJCAHO audit score. Financial performance was added because it is important forthe survival of hospitals. Survey respondents recorded their hospital’s relative per-formance over the last 3 years (see Appendix). Obtaining performance informationon a primary basis has extensive precedence in the literature (Powell, 1995). Thefive-item measure used in this study was adapted from scales used by Powell (1995)and Ramanujam, Venkatraman, and Camillus (1986).

The second added outcome measure was the hospital’s overall JCAHO auditscore. This industry expert-rated measure is based on a comprehensive audit offorty-nine standards related to internal hospital processes conducted once every3 years. The maximum JCAHO overall score is 100 points, and hospitals are underconsiderable institutional pressure to perform well on this audit (Westphal et al.,1997).

Our measure of customer satisfaction combines three dimensions of the hos-pital’s perceived performance relative to competitors, patient satisfaction, clinicaloutcomes, and average length of stay. Superior performance with respect to all threeof these dimensions should allow the hospital to meet or exceed the expectationsof their customers.

Two control variables were also measured: the number of hospital beds and thetype of ownership. The number of hospital beds is a proxy for organizational size.This measure is similar to those used in other studies of hospitals (e.g., Ketchen,Thomas, & Snow, 1993). The number of hospital beds was obtained from the 1995American Hospital Association’s Annual Survey of Hospitals. Hospital ownershipis a potentially important variable in this industry. For-profit hospitals may havedifferent organizational goals and unique groups of stakeholders when compared totheir not-for-profit counterparts (Zajac & Shortell, 1989). These data were collectedfrom the 1996 edition of the Hospital Blue Book.

Page 12: Deming Model

404 Evaluating the Deming Management Model

MBNQA Enhancements to the Deming Management Model

Since the first MBNQA awards ceremony in 1988, few research studies have beenconducted to evaluate the underlying framework of the MBNQA framework untilrecently (Flynn & Saladin, 2001). The criteria underlying the awards have evolvedover time. While the names of the categories have remained fairly constant, theirdefinitions and weights have changed somewhat in the process. In addition, anaward specific to health care organizations was introduced in 1995 (Meyer &Collier, 2001) and first bestowed in 2002.

Comparing the Deming management model to the general model depicted bythe MBNQA framework, Wilson and Collier (2000) argued that the MBNQA modelis more comprehensive and depicts many more causal paths. Since both modelsfocus on leadership, process information and management, customers, employees,and results, the recent studies evaluating the MBNQA may contribute importantinsights relevant to the Deming management model. Given the comprehensiveconstruction of the MBNQA framework (Bell & Keys, 1998), enhancements to theDeming management model based on these insights add to its generalizability.

Meyer and Collier (2001) and Wilson and Collier (2000) argue that the basicconcept underlying the MBNQA model is that “leadership drives the system thatcreates results.” Both of these studies, as well as the one by Flynn and Saladin(2001), confirmed this proposition. While the Deming management model alsobegins with leadership, paths from this construct only lead to internal and exter-nal cooperation and learning. The conceptual foundation supporting the MBNQAcriteria and the results from the initial MBNQA studies suggest that additionalpaths from leadership to other intermediate constructs in the Deming managementmodel exist. This suggestion is also supported by Anderson et al. (1995) in theirdiscussion of large unexplained effects in their path model and their suggestionthat the possibility of additional paths emanating from leadership may exist.

The MBNQA research also suggests the potential existence of direct pathsfrom leadership to business results and customer satisfaction. While the MBNQAframework contains these direct paths, Wilson and Collier (2000) did not find sig-nificant direct relationships. However, Meyer and Collier (2001), using a sample ofhospitals, did find a significant direct relationship between leadership and organi-zational performance. In addition, Flynn and Saladin (2001), using the world-classmanufacturing database, found significant relationships between leadership andboth business results and customer satisfaction. Therefore, there exists some evi-dence from these studies that leadership may be directly related to performance.

Finally, all three of these studies found significant relationships betweenprocess management and business or organizational results. While the Demingmanagement model does not contain this direct path, Anderson et al. (1994a)suggested that a direct path may exist between process management and customersatisfaction. Given that we are using a different database, we will perform an adhoc test of the paths suggested by these research studies.

SAMPLE CONSTRUCTION AND DATA COLLECTION

The study was conducted within the General Medical Hospitals (SIC 8062) indus-try. TQM has been recommended to the members of this industry as a strategy that

Page 13: Deming Model

Douglas and Fredendall 405

will assist them in dealing with their turbulent environment (Gaucher & Coffey,1993). Thus, this context was expected to provide an excellent platform on whichto test the subject model.

Data for the analysis were gathered on both a primary and secondary basis.In 1996 we randomly selected 19 metropolitan areas across the United States andsent questionnaires to the CEO and the Director of Quality in each hospital inthose Standard Metropolitan Statistical Areas (SMSA). Finally, we combined thesurvey data with secondary information available for the responding hospitals. Onehundred ninety-three out of the 512 hospitals in the sampling frame responded withat least one questionnaire, resulting in an overall response rate of 38 percent.

In an effort to assess the potential for response bias, comparisons were madeacross a number of available variables for both the responding and the nonre-sponding hospitals for the 19 SMSAs using data from the 1995 American HospitalAssociation Survey. The mean results of the nonrespondents did not differ signif-icantly from the responding firms when comparing assets, number of employees,profitability, or services offered. As a result, there does not appear to be systematicresponse bias in the financial and operating characteristics of the hospitals sampled.Although the database and a subset of the scales were used in a previously publishedstudy (Douglas & Judge, 2001), the conceptual framework, issues addressed, andmethodology are unique to this study.

METHODS

We chose to use structural equation modeling to estimate the relationships in thepath diagram displayed in Figure 1. LISREL 8.5 was the analytical procedure usedto estimate this model. This technique combines path analysis with multiple regres-sion analysis (Joreskog & Sorbom, 1993) in a manner that matches the theoreticalmodel displayed in Figure 1. The chi-square test associated with this model is χ2 =2,466.7, with 1,466 degrees of freedom (p = .001). The fit of the model was testedusing the Comparative Fit Index (CFI) suggested by Bentler (1990). While manyfit indices have been developed, the CFI was recommended in a review and evalua-tion of such indices by Medsker, Williams, and Holahan (1994). Values of the CFIshould realistically range from 0 to 1, with the values closest to 1 representing thebest fit (Marsh, Balla, & McDonald, 1988). The value of the CFI calculated in thisstudy was .86, suggesting that the model estimated fits the data sufficiently well.

RESULTS

The results will be presented for each hypothesis. Because multiple scales wereused to represent many of the constructs, we will present the results related to eachmeasure. Table 3 contains the results of the structural equation model estimatedusing LISREL.

Hypothesis 1

Our first hypothesis suggested that visionary leadership was related to higher lev-els of internal and external cooperation. We used two measures to represent in-ternal and external cooperation, quality philosophy and supplier cooperation. Top

Page 14: Deming Model

406 Evaluating the Deming Management Model

Tabl

e3:

Res

ults

from

the

stru

ctur

alm

odel

.

Dep

ende

ntV

aria

bles

Exp

lana

tory

Qua

lSu

pplr

TQ

Cus

tT

QC

ont

Stru

cC

ust

Fina

ncl

JCA

HO

Var

iabl

esPh

ilC

oop

Tra

inin

gD

rive

nM

BF

Met

hods

Impr

vmt

Exp

lor

Satis

fctn

Perf

Scor

e

TM

TIn

volv

emen

t.6

9∗∗∗

.49∗∗

∗.7

8∗∗∗

.57∗∗

∗Q

ualP

hil

.44∗∗

∗.2

3∗Su

pplr

Coo

p.0

0.0

2T

QT

rain

ing

.05

.14∗∗

Cus

tDri

ven

.48∗∗

∗.7

3∗∗∗

MB

F.2

9∗∗∗

.66∗∗

∗T

QM

etho

ds.3

7∗∗∗

.32∗∗

Con

tIm

prvm

t.2

2.1

91.

30St

ruc

Exp

lor

.50∗∗

∗.3

2∗∗.3

3R

2.7

8.4

0.5

3.6

4.6

4.7

0.6

0.5

2.3

0.1

3.0

5

∗∗∗ p

<.0

01;∗∗

p<

.01;

∗ p<

.05;

n=

193.

Page 15: Deming Model

Douglas and Fredendall 407

management team involvement, our measure of visionary leadership, was signif-icantly related to both quality philosophy (t = 10.80, p < .001) and supplierinvolvement (t = 7.59, p < .001). Therefore, Hypothesis 1 is supported.

Hypothesis 2

Our second hypothesis, that visionary leadership is related to higher levels of learn-ing, was tested using two measures of learning, total quality training and customer-driven information. Top management team involvement was significantly relatedto total quality training (t = 9.02, p < .001) and customer-driven information(t = 9.18, p < .001), indicating support for Hypothesis 2.

Hypothesis 3

Hypothesis 3 suggested that internal and external cooperation was related to pro-cess management. In addition to the two measures used for internal and externalcooperation, two measures were also used for process management, managementby fact and total quality methods. Quality philosophy was significantly related toboth management by fact (t = 3.98, p < .001) and total quality methods (t = 2.03,p < .05). However, our second measure of internal and external cooperation, sup-plier cooperation, was not related to either of the process management measures.This indicates moderate support for Hypothesis 3.

Hypothesis 4

The fourth hypothesis suggested that learning was positively related to processmanagement. Our results indicated that of the two measures representing learn-ing, only customer-driven information (4.35, p < .001) is significantly related tomanagement by fact. However, both total quality training (t = 2.15, p < .01) andcustomer-driven information (t = 5.69, p < .001) are significantly related to totalquality methods. Thus, we found general support for Hypothesis 4.

Hypothesis 5

Hypothesis 5 suggested that process management was positively related to con-tinuous improvement. Both management by fact (t = 4.42, p < .001) and totalquality methods (t = 5.70, p < .001) were significantly related to continuousimprovement. These results provide strong support for Hypothesis 5.

Hypothesis 6

The sixth hypothesis stated that process management is positively related to em-ployee fulfillment. Structural exploration measures employee fulfillment in ourstudy. Both management by fact (t = 4.37, p < .001) and total quality methods(t = 2.41, p < .01) were significantly related to employee fulfillment. Thus, supportwas found for Hypothesis 6.

Hypothesis 7

Hypothesis 7 suggested that continuous improvement was positively related tobusiness performance. The Deming management model includes customer satis-

Page 16: Deming Model

408 Evaluating the Deming Management Model

Table 4: Results from the cash flow margin regressionanalysis.

Independent Variables β T

Hospital Beds −.02 −.24Ownership −.45 −5.47∗∗Continuous Improvement .17 2.12∗

Adjusted R2 .20F 11.21∗∗

∗p < .05; ∗∗p < .01; n = 122.

faction as a performance variable. We have added financial performance and eachhospital’s JCAHO audit score as additional measures of business performance. Inour study, we did not find significant relationships between continuous improve-ment and financial performance or customer satisfaction. With respect to the JC-AHO audit score (t = 1.79, p < .10), marginal significance was found. Therefore,Hypothesis 7 was generally not supported.

We were able to conduct one additional test of the relationship betweencontinuous improvement and financial performance. Cash flow margin data wereavailable archivally for 122 of the 193 hospitals in the sample. We verified the repre-sentativeness of this subsample by comparing key variables (size, services, JCAHOscore, perceived financial performance) across the remaining hospitals versus thosethat dropped out. Cash flow margin is a financial performance variable often usedin empirical studies of hospitals (Douglas & Ryman, 2003). Using the averagecash flow margin for the years 1996–1997, we conducted a regression analysisusing continuous improvement as the independent variable and included the samecontrol variables mentioned above. The results displayed in Table 4 demonstratethat continuous improvement is significantly related to cash flow margin (t = 2.12,p < .05) using the objective data added to our database.

Hypothesis 8

Our final hypothesis suggested that employee fulfillment was positively related tobusiness performance. We found the structural exploration measure to be signif-icantly related to both customer satisfaction (t = 2.99, p < .001) and financialperformance (t = 2.49, p < .01). However, it was not significant with respect tothe JCAHO audit score. Thus, Hypothesis 8 is supported in our study for two ofthe key measures of business performance.

Finally, neither of the control variables was significant in our study. Neithersize nor ownership type of the hospitals in this database was significantly relatedto business performance.

Ad Hoc Tests Suggested by the MBNQA Literature

Results of the ad hoc tests are contained in Table 5. Starting with the Deming man-agement model results, the LISREL program was used to test the addition of thepaths suggested by the MBNQA studies on a nested basis. Figure 2 overlays the

Page 17: Deming Model

Douglas and Fredendall 409

Table 5: Ad hoc tests suggested by MBNQA studies.

CriticalSteps df �df χ2 �χ2 Value p

Deming Mgt Model 1,466 2,466.73Step 1: Leadership Paths to

Continuous Improvement,Process Management, andEmployee Fulfillment

1,462 4 2,435.95 31.23 18.47 <.001

Step 2: Leadership Paths toFinancial Performance

1,461 1 2,433.86 2.09 3.84 N.S.

Customer Satisfaction 1,460 1 2,425.03 8.83 6.63 <.01Step 3: Process Management

Paths to1,458 2 2,422.67 2.36 5.99 N.S.

Financial PerformanceCustomer Satisfaction

1,456 2 2,416.71 5.96 5.99 N.S.

Figure 2: Revised theory of quality management for services.

Organizational

System

Process

Outcomes

Internal and

External

Cooperation

Continuous

Improvement

Visionary

Leadership

Process

Management

Business

Performance

Learning Employee

Fulfillment

Note: The additional paths suggested by the MBNQA literature that were significant wereincluded as follows:

Step 1 is represented byStep 2 is represented by

significant paths onto the conceptual Deming management model. Our first stepwas to add the additional paths between leadership and the remaining intermediateconstructs, process management, continuous improvement, and employee fulfill-ment. This step added four additional paths to the model and resulted in a changein χ2 of 31.23, which is significant (p < .001). This revised model, therefore,represents a better fit with this database. The second step added paths betweenleadership and two of the performance measures, customer satisfaction and finan-cial performance. Each path was added individually, and only adding the path fromleadership to customer satisfaction resulted in an improved model fit with a change

Page 18: Deming Model

410 Evaluating the Deming Management Model

in χ2 of 8.83 (p < .01). Finally, the third step added paths from process manage-ment to customer satisfaction and financial performance. Adding these paths didnot result in a better fitting model.

DISCUSSION

The main focus of this research has been a conceptual and empirical evaluation ofquality management in the context of a service industry, initially using the Demingmanagement model. This model had been previously tested only in manufacturingindustries. In addition, changes to the model, suggested in recent studies focusedon the more general MBNQA framework, were tested and found to be importantenhancements.

This study represents an additional replication of the Deming managementmodel and contributes to the discussion concerning its generalizability. Rungtu-sanatham et al. (1998) recommended that structural equation modeling be usedwith samples from additional U.S. industries before reaching conclusions regard-ing the utility of the model. With this in mind, we will compare our findings withthose of the two earlier studies.

Hypotheses 1, 2, 5, and 6 were fully supported. Some support was identifiedfor Hypotheses 3, 4, and 8 and none for Hypothesis 7 from our LISREL analysis.These results are similar to those found by the two earlier studies in manufacturingindustries, with a few notable exceptions. First, we used multiple measures to rep-resent a number of the model concepts. This should allow us to better understandthe relationships central to the model. Second, we found support for the relation-ship between learning and process management, which each of the earlier studiesfailed to do. Third, we added two additional performance variables as suggested bythe MBNQA literature—financial performance and the JCAHO audit score—inorder to better understand the breadth of the outcomes related to implementingquality management. These additional measures are necessary to fully assess theeffectiveness of TQM (Hackman & Wageman, 1995).

Similarly to the manufacturing studies, we found support for the relationshipbetween internal and external cooperation and process management. However,rather than combining both types of cooperation into one measure, we used sep-arate measures. In our study, the presence of a quality philosophy representedinternal cooperation. We found that this measure was significantly related to pro-cess management. However, our external measure, supplier cooperation, was notrelated to process management. It may be that, in this industry, close cooperationwith suppliers has little relevance to hospital processes. Or, it is possible that therespondents to the survey were not in a position to evaluate the level of cooperationwith suppliers. The survey respondents were located at individual hospitals, whilemany of the supplier contacts were made at a central location in the hospital’scorporate structure.

In this study, learning was represented with two measures, total quality train-ing and customer-driven information. We also included two measures of processmanagement—management by fact and total quality methods. All relationshipswere significant except the one between total quality training and managementby fact. Neither Anderson et al. (1995) nor Rungtusanatham et al. (1998) found a

Page 19: Deming Model

Douglas and Fredendall 411

significant relationship between learning and process management. The significantrelationship between learning and process management is important. While it hasbeen argued by many that learning is a key concept within quality management(Anderson et al., 1994a; Hackman & Wageman, 1995; Huq & Martin, 2000), littlesupportive evidence has existed.

The path in the model from process management to employee fulfillment wassupported. There were positive relationships between both process managementmeasures and employee fulfillment measures. Since exploration reflects the abilityof organizational members to use readily available information for decision making,this significant path indicates that supplying and utilizing quality data is critical toemployee decision making.

The reason that Anderson et al. (1995) and Rungtusanatham et al. (1998) didnot find a significant relationship between process management and employee ful-fillment may be due to the operationalization of the employee fulfillment variable.Anderson et al. (1994a, p. 480) defined employee fulfillment as “exemplified byjob satisfaction, job commitment, and pride of workmanship.” However, Andersonet al. (1995) and Rungtusanatham et al. (1998) used three and two items, respec-tively, which seemed to measure only pride of workmanship.

The structural exploration items in this study measured the level of informa-tion sharing and employee adaptation, which are expected to result in job satis-faction and commitment (Robbins et al., 2002). The lack of support for employeefulfillment in earlier studies may be attributable to the lack of a metric that cor-relates well with job satisfaction or job commitment. Future work should exploremore complete operationalizations of this key construct.

With respect to the relationship between employee fulfillment and businessperformance, we found significant relationships between structural explorationand both financial performance and customer satisfaction. Anderson et al. (1995)also found a strong relationship between employee fulfillment and customer sat-isfaction. Rungtusanatham et al. (1998) did not, but attributed the problem to theculture in the Italian manufacturing plants in their study. Our finding with re-spect to structural exploration suggests additional studies may be needed to betterunderstand or even revise the role of employee involvement in quality implemen-tation, especially in service organizations. Higher levels of exploration were re-lated to higher levels of customer satisfaction and financial performance relative tocompetitors.

In the manufacturing studies, Anderson et al. (1995) found no relationship be-tween continuous improvement and customer satisfaction while Rungtusanathamet al. (1998) found marginal significance. Our study did not find a significantrelationship between continuous improvement and any of the three performancemeasures. This is surprising, because there is strong theoretical support for this re-lationship. One possible explanation is that the hospitals’ continuous improvementefforts may be focused on cost control or cost reductions, which would not impactcustomer satisfaction directly. Our additional test using the subset of the sample forwhich objective financial results were available did find a significant relationshipbetween continuous improvement and financial performance. This ad hoc findingsuggests that a relationship may exist and that further research is needed to betterinvestigate this relationship and the relevant constructs.

Page 20: Deming Model

412 Evaluating the Deming Management Model

In addition, there are two other possible explanations suggested by the lit-erature for the lack of a significant relationship in this study between continuousimprovement and customer satisfaction. First, neither this study nor the prior twostudies of the Deming management model directly measured customer satisfac-tion. Second, it is possible that service quality is an independent construct thatis an antecedent to customer satisfaction. Both explanations are involved in thedebate about the appropriate use of instruments such as SERVQUAL to measurecustomer satisfaction (Van Dyke, Kappelman, & Prybutok, 1997) and are discussedin greater detail below.

The ad hoc tests based on the MBNQA literature are an important contributionto the literature. The additional four paths from leadership to the other variablesshown in Figure 2 significantly increased the model fit and suggest that, as originallyconceptualized, the Deming management model may be too parsimonious. Thesignificance of these paths does support Deming’s stress on the importance ofleadership to business improvement. These ad hoc tests suggest that leadership hasa direct effect on every subsequent variable in the model, strongly supporting theunderlying concepts associated with the MBNQA model. Because the derivationof the MBNQA model has a broader base than the initial focus on Deming’s worktaken by Anderson et al. (1994a), the enhancements suggested for the Demingmanagement model may prove more useful going forward.

While the founders of total quality stressed its application to services, mostservice research has ignored many of the elements of the quality managementmodel. The significant findings of this research are an important contribution to thequality and service literatures in general and to health services in particular. Thesefindings suggest that it is important that service quality researchers integrate moreelements of quality management into their research.

In addition, it is important that future research incorporate insights fromthe service quality literature into the Deming management model. One of the keyquestions for investigation is whether service quality is a distinct variable precedingcustomer satisfaction as suggested by Dabholkar, Shepherd, and Thorpe (2000),or is service quality already contained in the customer satisfaction construct? Asecond, related question is what instrument is most appropriate to use to measurecustomer satisfaction?

Dabholkar et al. (2000) viewed customer satisfaction as an overall judgmentof service quality and suggested that it be measured separately from service quality.They defined service quality as a set of dimensions such as reliability and respon-siveness that preceded customer satisfaction. They found the service quality andcustomer satisfaction constructs highly correlated, but distinct in the retail indus-try. These issues have not yet been examined in great depth outside of the retailindustry.

SERVQUAL is a frequently used instrument measuring customer satisfac-tion in the service quality literature (Harvey, 1998; Kettinger & Lee, 1997). Thedefinition of service quality underlying SERVQUAL (Parasuraman, Zeithaml, &Berry, 1988) is equivalent to the definition of customer satisfaction associated withthe Deming management model that is displayed in Table 1. Therefore, it would beappropriate to use SERVQUAL to measure whether customers perceive that their

Page 21: Deming Model

Douglas and Fredendall 413

needs are being met. Many researchers support the continued use of SERVQUAL tomeasure customer satisfaction, but recommend that more work be done to improveits scales (Kettinger & Lee, 1999).

The suggestion by Dabholkar et al. (2000) presented above, that servicequality is conceptually different from customer satisfaction, is based on a differentdefinition of service quality. According to Voss, Roth, Rosenzweig, Blackmon, andChase (2004), service quality, as discussed in the operations management literature,depends on meeting or exceeding service standards, which are established withinthe provider organization. The question of whether a distinct, internal measure ofservice quality exists and whether it should be added to the Deming managementmodel is an important topic for future research.

In summary, two important issues that need to be resolved are whether servicequality, as defined in the operations management literature, should be included inthe Deming management model as a separate construct, and what is the mostappropriate method to use to measure customer satisfaction? Given reliable andvalid measures of service quality, if it is a necessary and distinct construct, andcustomer satisfaction, their use in the enhanced Deming management model maylead to a better understanding of the outcomes of quality management for bothservices and manufacturing.

Implications for Managers

The findings from this study should also be useful to managers and administratorsinterested in a continued focus on quality improvement. Since TQM/CQI has notbeen consistently adopted across hospitals (Westphal et al., 1997) and the prac-tices underlying these concepts are embedded in more recent programs such asReengineering, Six Sigma, and Lean Production, the implications of our findingsmay help focus managerial attention on a more complete implementation of therelevant management model. It is also worthy to note the importance of includingthe MBNQA enhancements, since the MBNQA model has been used as a modelfor establishing awards in many of the individual states in the United States andin countries abroad, including the European Quality Award and those in Brazil,Egypt, Japan, and Mexico (Flynn & Saladin, 2001).

The importance of leadership, emphasized in the findings with respect tothe initial paths to cooperation and learning and the secondary paths suggested bythe MBNQA, should be of paramount importance for hospital administrators andmanagers within other service industries. This finding also directly addresses a keyissue raised by Arndt and Bigelow (1995). These authors questioned the abilityof the CEO and other top administrators to adequately provide the needed visionand guidance for a quality implementation in a hospital environment. Our findingssuggest that these managers can and do have a significant impact on all aspects ofa quality system and need to continue to take an active part to ensure success.

Our study also provides empirical evidence that total quality training andthe availability and use of customer-driven information are closely related to im-proved process management. This finding supports efforts of hospital managersto develop a learning culture. As Bender and Krasnick (1995) have pointed out,

Page 22: Deming Model

414 Evaluating the Deming Management Model

few hospitals exhibit such a culture and yet developing one is important for a totalquality implementation in this industry.

The finding that process management is significantly related to continu-ous improvement suggests that managers in service industries need to empha-size the use of basic process management tools such as management by factand total quality management methods to enhance their continuous improvementefforts. As pointed out by Hackman and Wageman (1995), many TQM organi-zations have focused little attention on using these scientific methods in theirprograms.

Overall, managers in service organizations should be aware that each of theconstructs underlying the Deming management model is important. In addition,leadership from the top of the organization is critically related to attaining higherlevels in each area. Limited managerial attention to the quality system jeopardizesa successful implementation of TQM.

Limitations

This study shares some of the same limitations as the Anderson et al. (1995)and Rungtusanatham et al. (1998) studies. The most significant limitation is thatthe scales used were not originally developed for the constructs in the Demingmanagement model. As stated earlier, this may be a partial explanation of whysome paths in the model are not significant or are only weakly supported. However,where possible we used multiple measures for most constructs in order to broadenour understanding of the relationships within the model. In addition, the measureswere obtained from the management literature where they had been evaluated forreliability and validity. It would be desirable for future studies to evaluate themeasures used here, the ones from the manufacturing studies, those associatedwith SERVQUAL, and others that may be appropriate to further our understandingof the Deming management model.

One advantage of this study, that all respondents were from the same industry,also limits the generalizability of the findings to other service industries. Futureresearch that applies the enhanced Deming management model to other industriesis needed. As Reeves and Bednar (1994) suggested that a concept as complex asservice quality management can only be understood through cumulative analysesof multiple industries.

In addition, while the causality implied by the model is supported by thetheory, the study design does not allow us to directly test the proposed causalitybecause the data were gathered at a point in time. Future studies of a longitudinalnature would be helpful to address this issue.

CONCLUSIONS

An important contribution of this research is the extension of the Deming manage-ment model to services. The strong theoretical and empirical support for many ofthe relationships in the Deming management model tested within the health careindustry suggests that it is important for researchers to incorporate these variables

Page 23: Deming Model

Douglas and Fredendall 415

into their research about service quality. The findings here suggest that the Demingmanagement model is as applicable to services as it is to manufacturing.

An additional contribution relates to the enhancements to the Demingmanagement model suggested by the development and testing of the MBNQAframework. The MBNQA criteria were established on a broader basis than theDeming model and better represent the quality implementations in organizations.The importance of having visionary leadership directly involved in all facets ofquality programs proved important in this dataset and represents a key message formanagers and scholars alike.

Significant questions remain unanswered with respect to performance mea-sures within the enhanced model. While conceptual arguments are strong withrespect to hypothesized relationships between continuous improvement and out-come variables representing customer satisfaction and operational and financialperformance, only marginal evidence was found to support the arguments. Thehope is that the findings here will encourage research integrating the rich servicequality literature into the model. Future research should investigate whether cus-tomer satisfaction mediates the relationship between service quality and businessperformance. As stated earlier this will require research into the distinction be-tween the service quality and customer satisfaction constructs. It will also requirea determination of whether it is better to measure customer satisfaction by directlyusing an instrument such as SERVQUAL.

Quality management continues to be an important research topic, especiallygiven its embeddedness in current managerial programs such as Six Sigma andLean Production. Use of the enhanced Deming management model, combinedwith the customer satisfaction work contained in the service quality literature, infuture research may lead to better-informed literatures focusing on both manu-facturing and service performance. [Received: February 2003. Accepted: January2004.]

REFERENCES

Anderson, E. W., Fornell, C., & Lehmann, D. R. (1994a). Customer satisfaction,market share, and profitability: Findings from Sweden. Journal of Marketing,58(July), 53–66.

Anderson, J., Rungtusanatham, M., & Schroeder, R. (1994b). A theory of qual-ity management underlying the Deming management method. Academy ofManagement Review, 19, 472–509.

Anderson, J., Rungtusanatham, M., Schroeder, R., & Devaraj, S. (1995). A pathanalytic model of a theory of quality management underlying the Demingmanagement method: Preliminary empirical findings. Decision Sciences, 26,637–658.

Arndt, M., & Bigelow, B. (1995). The implementation of total quality managementin hospitals: How good is the fit? Health Care Management Review, 20(4),7–14.

Page 24: Deming Model

416 Evaluating the Deming Management Model

Bell, R., & Keys, B. (1998). A conversation with Curt W. Reimann on the back-ground and future of the Baldrige award. Organizational Dynamics, 26(4),51–61.

Bender, A., & Krasnick, C. (1995). Total quality management: Lessons learnedfrom the medical practice environment. Health Care Supervisor, 13(4), 70–76.

Bentler, P. (1990). Comparative fit indexes in structural models. PsychologicalBulletin, 107, 237–246.

Boerstler, H., Foster, R., O’Connor, E., O’Brien, J., Shortell, S., Carmen, J., &Hughes, E. (1996). Implementation of total quality management: Conven-tional wisdom versus reality. Hospital & Health Services Administration,41(2), 143–159.

Bureau of Labor Statistics. (2002). Current employment statistics survey (Na-tional). Washington, DC: U.S. Department of Labor.

Callahan, J., & Moretton, B. (2001). Reducing software product development time.International Journal of Project Management, 19, 59–70.

Carmen, J., Shortell, S., Foster, R., Hughes, E., Boerstler, H., O’Brien, J., &O’Connor, E. (1996). Keys for successful implementation of total qualitymanagement in hospitals. Heath Care Management Review, 21(1), 48–60.

Chase, R. B. (1996). The mall is my factory: Reflections of a service junkie.Production and Operations Management, 5(4), 298–308.

Cook, L. S., & Verma, R. (2002). Exploring the linkages between quality system,service quality, and performance excellence: Service providers’ perspectives.Quality Management Journal, 9(2), 44–56.

Dabholkar, P., Shepherd, C. D., & Thorpe, D. I. (2000). A comprehensive frame-work for service quality: An investigation of critical conceptual and mea-surement issues through a longitudinal study. Journal of Retailing, 76(2),139–173.

Deming, W. (1986). Out of the crisis. Cambridge: Massachusetts Institute of Tech-nology Press.

Douglas, T., & Judge, W. (2001). Total quality management implementationand competitive advantage: The role of structural control and exploration.Academy of Management Journal, 44, 158–169.

Douglas, T., & Ryman, J. (2003). Understanding competitive advantage in thegeneral hospital industry: Evaluating strategic competencies. Strategic Man-agement Journal, 24, 333–347.

Flynn, B. B., Schroeder, R. G., & Sakakibara, S. (1994). A framework for qualitymanagement research and an associated measurement instrument. Journal ofOperations Management, 11, 339–366.

Flynn, B. B., & Saladin, B. (2001). Further evidence on the validity of the theoreticalmodels underlying the Baldrige criteria. Journal of Operations Management,19, 617–652.

Page 25: Deming Model

Douglas and Fredendall 417

Foster, S. T., Howard, L. W., & Shannon, P. (2002). The role of quality tools inimproving satisfaction with government. Quality Management Journal, 9(3),20–31.

Gaucher, E., & Coffey, R. (1993). Total quality in healthcare. San Francisco:Jossey-Bass.

Ghobadian, A., Speller, S., & Jones, M. (1994). Service quality: Concepts andmodels. International Journal of Quality and Reliability Management, 11(9),43–66.

Gittell, J. H. (2002). Relationships between service providers and their impact oncustomers. Journal of Service Research, 4(4), 299–311.

Goal/QPC. (1991). The Memory Jogger Plus+ Videotape Series. Methuen, MA:Goal/QPC with Amatulli & Associates, Inc.

Hackman, J., & Wageman, R. (1995). Total quality management: Empirical, con-ceptual, and practical issues. Administrative Science Quarterly, 40, 309–342.

Harvey, J. (1998). Service quality: A tutorial. Journal of Operations Management,16, 583–597.

Hensel, J. S. (1990). Service quality improvement and control: A customer-basedapproach. Journal of Business Research, 20, 43–54.

Huq, Z., & Martin, T. (2000). Workforce cultural factors in TQM/CQI implemen-tation in hospitals. Health Care Management Review, 25(3), 80–103.

Joreskog, K., & Sorbom, D. (1993). LISREL 8: Structural equation modeling withthe SIMPLIS command language. Chicago: Scientific Software International,Inc.

Kellogg, D. L., & Chase, R. B. (1995). Constructing an empirically derived measurefor customer contact. Management Science, 41(11), 1734–1749.

Ketchen, D., Thomas, J., & Snow, C. (1993). Organizational configurations andperformance: A comparison of theoretical approaches. Academy of Manage-ment Journal, 36, 1278–1313.

Kettinger, W. J., & Lee, C. C. (1997). Pragmatic perspectives on the measurementof information systems service quality. MIS Quarterly, 21(2), 223–240.

Kettinger, W. J., & Lee, C. C. (1999). Replication of measures in informationsystems research: The case of IS SERVQUAL. Decision Sciences, 30, 893–899.

Khandwalla, P. (1977). The design of organizations. New York: Harcourt BraceJovanovich, Inc.

Kunst, P., & Lemmink, J. (2000). Quality management and business performancein hospitals: A search for success parameters. Total Quality Management, 11,1123–1133.

Marsh, H., Balla, J., & McDonald, R. (1988). Goodness-of-fit indexes in confirma-tory factor analysis: The effect of sample size. Psychological Bulletin, 103,391–410.

Page 26: Deming Model

418 Evaluating the Deming Management Model

Medsker, G., Williams, L., & Holahan, P. (1994). A review of current practicesfor evaluating causal models in organizational behavior and human resourcesmanagement research. Journal of Management, 20, 439–464.

Meyer, S., & Collier, D. (2001). An empirical test of the causal relationships in theBaldrige health care pilot criteria. Journal of Operations Management, 19,403–425.

Motwani, J., Sower, V., & Brashier, L. (1996). Implementing TQM in the healthcare sector. Health Care Management Review, 21(1), 73–82.

National Institute of Standards and Technology (NIST). (2003). Malcolm BaldrigeNational Quality Program Health Care Criteria for Program Excellence.Gaithersburg, MD.

Nilsson, L., Johnson, M. D., & Gustafsson, A. (2001). The impact of quality prac-tices on customer satisfaction and business results: Product versus serviceorganizations. Journal of Quality Management, 6, 5–27.

Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1985). A conceptual model ofservice quality and its implications for future research. Journal of Marketing,49(Fall), 41–50.

Parasuraman, A., Zeithaml, V. A., & Berry, L. L. (1988). SERVQUAL: A multiple-item scale for measuring consumer perceptions of service quality. Journal ofRetailing, 64(1), 12–37.

Powell, T. (1995). Total quality management as competitive advantage: A reviewand empirical study. Strategic Management Journal, 16, 15–37.

Ramanujam, V., Venkatraman, N., & Camillus, J. (1986). Multi-objective assess-ment of effectiveness of strategic planning. Academy of Management Journal,29, 347–372.

Reeves, C., & Bednar, D. (1994). Defining quality: Alternatives and implications.Academy of Management Review, 19, 419–445.

Robbins, T. L., Crino, M. D., & Fredendall, L. D. (2002). An integrative modelof the empowerment process. Human Resource Management Review, 12,419–433.

Roth, A. V., & Jackson, W. E. III (1995). Strategic determinants of service qualityand performance: Evidence from the banking industry. Management Science,41(11), 1720–1733.

Routhieaux, R., & Gutek, B. (1998). TQM/CQI effectiveness at team and depart-ment levels. Journal of Quality Management, 3, 39–62.

Rungtusanatham, M., Forza, C., Filippini, R., & Anderson, J. (1998). A replicationstudy of a theory of quality management underlying the Deming managementmethod: Insights from an Italian context. Journal of Operations Management,17, 77–95.

Rust, R., Zahorik, A. J., & Keiningham, T. L. (1995). Return on quality (ROQ):Making service quality financially accountable. Journal of Marketing, 59(2),58–70.

Page 27: Deming Model

Douglas and Fredendall 419

Saraph, J., Benson, P., & Schroeder, R. (1989). An instrument for measuringthe critical factors of quality management. Decision Sciences, 20, 810–829.

Shortell, S., O’Brien, J., Carman, J., Foster, R., Hughes, E., Boerstler, H., &O’Conner, E. (1995). Assessing the impact of continuous quality improve-ment/total quality management: Concept versus implementation. Health Ser-vices Research, 30, 377–401.

Shostack, G. L., (1984). Designing services that deliver. Harvard Business Review(January–February), 133–139.

Sousa, R., & Voss, C. A. (2002). Quality management re-visited: A reflectivereview and agenda for future research. Journal of Operations Management,20, 91–109.

Stewart, D. M., & Chase, R. B. (1999). The impact of human error on deliv-ering service quality. Production and Operations Management, 8(3), 240–263.

Sureshchandar, G. S., Rajendran, C., & Anantharaman, R. N. (2001). A conceptualmodel for total quality management in service organizations. Total QualityManagement, 12(3), 343–363.

Waldman, D. (1994). The contributions of total quality management to a theory ofwork performance. Academy of Management Review, 19, 510–536.

Westphal, J., Gulati, R., & Shortell, S. (1997). Customization or conformity? Aninstitutional and network perspective on the content and consequences ofTQM adoption. Administrative Science Quarterly, 42, 366–394.

Wilson, D., & Collier, D. (2000). An empirical investigation of the MalcolmBaldrige National Quality Award Causal Model. Decision Sciences, 31, 361–383.

Van Dyke, T. P., Kappelman, L. A., & Prybutok, V. R. (1997). Measuring infor-mation systems service quality: Concerns on the use of the SERVQUALquestionnaire. MIS Quarterly, 21(2), 195–209.

Voss, C. A., Roth, A. V., Rosenzweig, E. D., Blackmon, K., & Chase, R. B.(2004). A Tale of Two Countries’ Conservatism, Service Quality, and Feed-back on Customer Satisfaction. Journal of Service Research, 6(3), 212–230.

Zabada, C., Rivers, P., & Munchus, G. (1998). Obstacles to the application of totalquality management in health-care organizations. Total Quality Management,9, 57–66.

Zajac, E., & Shortell, S. (1989). Changing generic strategies: Likelihood, direc-tion, and performance implications. Strategic Management Journal, 10, 413–430.

Zeitz, G., Johannesson, R., & Ritchie, J. (1997). An employee survey measuringtotal quality management practices and culture: Development and culture.Group & Organization Management, 22, 414–434.

Page 28: Deming Model

420 Evaluating the Deming Management Model

APPENDIX

Questionnaire Items Used in This Study‡

TQM Practices: Scale represents extent to which items are practiced in your orga-nization (1 = very low to 5 = very high)

1. Top Management Team Involvement• The top health care organization executive assumes responsibility for

quality performance.

• The major department heads participate in the quality improvementprocess.

• The organization’s top management (top administrator and majordepartment heads) has objectives for quality performance.

• The goal-setting process for quality within the health care organizationis comprehensive.

• Importance is attached to quality by the organization’s top managementin relation to cost objectives.

• Quality issues are reviewed in the organization’s top managementmeetings.

2. Quality Philosophy• There is a strong commitment to quality at all levels of this organization.

• People in this organization are aware of its overall mission.

• Members of this organization show concern for the need for quality.

• Continuous quality improvement is an important goal of thisorganization.

• Managers here try to plan ahead for changes that might affect ourperformance.

3. Total Quality Training• Quality-related training is given to hourly employees throughout the

organization.

• Quality-related training is given to managers and supervisors throughoutthe organization.

• Training is given in the “total quality concept” (i.e., philosophy ofcompany-wide responsibility for quality) throughout the organization.

• Training is given in the basic statistical techniques (such as histogramsand control charts) in the organization as a whole.

• The organization’s top management is committed to employee trainingfor quality.

• Resources are provided for employee training in quality.

4. Customer Driven Information• Associates know who their customers are.

• Associates attempt to measure their internal customers’ needs(customers inside this organization).

Page 29: Deming Model

Douglas and Fredendall 421

• Associates attempt to measure their external customers’ needs(customers outside this organization).

• The organization uses customer requirements as the basis for Quality.

• Our organization is more customer focused than our competitors.

5. Continuous Improvement• Associates in the organization try to improve the quality of their service.

• Associates in the organization believe that quality improvement is theirresponsibility.

• Associates in the organization analyze their work products to look forways of doing a better job.

6. Management by Fact• Quality data (defects, complaints, outcomes, time, satisfaction, etc.) are

available.

• Quality data are timely.

• Quality data are used as tools to manage quality.

• Quality data are available to hourly workers.

• Quality data are available to managers and supervisors.

• Quality data are used to evaluate supervisor and managerialperformance.

7. Total Quality Methods• Associates use the basic statistical techniques (such as histograms and

control charts) to study their work processes.

• Associates analyze the time it takes to get the job done.

• Associates keep records and charts measuring the quality of work dis-played in their work area.

• Statistical techniques are used to reduce variation in processes in theorganization.

• TQM procedures (such as brainstorming, cause-and-effect diagrams,Pareto charts) are used to analyze information for process improvement.

8. Supplier Involvement• Suppliers are selected based on quality rather than price.

• The organization’s supplier rating system is thorough.

• The organization relies on reasonably few, but dependable suppliers.

• The organization provides education to its suppliers.

• Longer term relationships are offered to suppliers.

• Clear specifications are provided to suppliers.

Perceived Financial Performance: Scale represents the organization’s relative per-formance to competitors over the last three years (1 = much worse, 5 = muchbetter).

� Growth in earnings� Growth in revenue

Page 30: Deming Model

422 Evaluating the Deming Management Model

� Changes in market share� Return on assets� Long-run level of profitability

Perceived Customer Satisfaction: Scale represents the organization’s relative per-formance to competitors over the last three years (1 = much worse, 5 = muchbetter).

� Patient satisfaction� Average length of stay� Clinical outcomes

Structural Exploration: Scale asks the respondent to identify the operating manage-ment philosophy actually used in their health care organization. A “1” representsthe expression on the left-hand side and a “7” represents the expression on theright-hand side, with “4” as the mid-point standing for a combination of the two.

Highly structured channels ofcommunication and highlyrestricted access toimportant financial andoperating information

1 2 3 4 5 6 7 Open channels of communicationwith important financial andoperating information flowingquite freely throughout theorganization

A strong emphasis on holdingfast to true and triedmanagement principlesdespite any changes inbusiness conditions

1 2 3 4 5 6 7 A strong emphasis on adaptingfreely to changingcircumstances without toomuch concern for pastpractices

‡A complete copy of the questionnaire can be obtained from the lead author.

Thomas J. Douglas is assistant professor of management at Clemson University.He received his PhD in strategic management from the University of Tennessee.He has more than 25 years of industry experience with SBC Communications.He has published or has forthcoming work in the Academy of Management Jour-nal, Strategic Management Journal, Journal of Business Venturing, and Interfaces,among others. His research interests are in the areas of competitive advantage, en-trepreneurship, total quality management, and sustainable environmental strategies.He is a member of the Academy of Management and the Strategic ManagementSociety.

Lawrence D. Fredendall is associate professor of management at Clemson Univer-sity. He received his MBA and his PhD in operations management from MichiganState University. His most recent book is titled Basics of Supply Chain Managementand was published by The St. Lucie Press/APICS Series. His work has appeared injournals such as the Journal of Operations Management, International Journal ofProduction Research, European Journal of Operational Research, and Productionand Operations Management.

Page 31: Deming Model