April-June 2020 Edition

Volume 20, Number 2 • April - June 2020 Sanjeev Verma and Integrated Framework for Vendor Management: Prateek Girdhar An Interpretivist Approach Jaume Franquesa, Explanation and Rigor in Management Theorizing: William Acar and Which Theory-Building Criteria Make for an Influential Jino O. Mwaka Contribution? Mala Srivastava and Impact of Service Failure Attributions on Dissatisfaction: Aditi Gosain Revisiting Attribution Theory Richard Brown and William Kline Optimal Misalignment: Strategic Intent, Organizational Capabilities, and Performance Quarterly Journal of Research in Management

Journal of Management Research Quarterly Journal of Research in Management CALL FOR PAPERS Journal of management Research is a refereed journal at par with the best international journals on the subject. The journal seeks manuscripts that identify, extend, unify, test or apply scientific and multi-disciplinary knowledge pertaining to the management field. Aims and Scope 6. The papers must be thoroughly vetted and finely read, and there must be no discrepancy regarding Journal of Management Research endeavours to spellings of the names of the authors and years of promote and disseminate knowledge in the complex publications from the ones mentioned in the text multi-disciplinary management field. The Journal of the paper with the citations given at the end encourages theoretical and empirical research papers of the paper. The citations must be complete in and articles of relevance to both academicians and every manner. practitioners. In addition, the journal invites papers covering application of theory to real life management 7. All the figures and illustrations must be in B&W, activities, where the findings would be of interest and in no case should be bigger than 6”x8”. to researchers, executives, academicians and management students. 8. Authors are responsible for making sure that they have not duplicated or copied, in full or in part, an The Journal publishes articles from areas such article/content already published. as finance, accounting, marketing, operations management, human resources management, 9. Authors should certify on the cover page of international business, information technology, the paper that the material is not published, environment, risk management, globalization and copyrighted, accepted or under review elsewhere. related areas. Articles reflecting diversity, cross- functional nature of management and emerging 10. The papers must be submitted only in electronic concerns are also considered for publication. form at [email protected]. Guidelines for Authors 11. Authors will be required to sign a Copyright Transfer Agreement after acceptance of the 1. The cover page of the paper should contain title paper. of the paper, author’s name, designation, official address, contact address, personal phone/mobile 12. Copyright of all accepted papers for publication numbers and personal and permanent e-mail ID. and use in any form/format/way, i.e., in text, image, other electronic format or such other 2. Papers should contain keywords, and a formats or on such other media as may now non-mathematical abstract of about 200 words. exist or hereafter be discovered, will vest with South Asia Publications. South Asia Publications 3. Papers should be of 3500-10000 words in length. is also authorized to market the contents of the journal to end-users through re-license of 4. Review of papers takes 2-4 weeks. sale of information products, using the media of CD-ROM, tape, on-line hosts, internet services 5. All notes and references should be placed at and other electronic or optical media or formats the end of the paper. The references list should now known or hereafter discovered. Permission mention only those sources which are cited in the to make digital/hard copy of published work is text. The citations should be typed in the format granted without fee for personal or classroom use. and style as specified by American Psychological Association. Please visit www.apastyle.org for The submissions should be sent to further details. [email protected]

Journal of Management Research Volume 20, Number 2 • April - June 2020 Contents 6 3 Integrated Framework for Vendor Management: An Interpretivist Approach Sanjeev Verma and Prateek Girdhar 75 Explanation and Rigor in Management Theorizing: Which Theory-Building Criteria Make for an Influential Contribution? Jaume Franquesa, William Acar and Jino O. Mwaka 99 Impact of Service Failure Attributions on Dissatisfaction: Revisiting Attribution Theory Mala Srivastava and Aditi Gosain 113 Optimal Misalignment: Strategic Intent, Organizational Capabilities, and Performance Richard Brown and William Kline Journal of Management Research, its editors and publisher disclaim responsibility and liability for any statement of facts and opinion, originality of contents, and of any copyright violations by the authors.

ISSN: 0972-5814 Editorial Board Online ISSN: 0974-455X Anbalagan Krishnan Volume 2, Number 2 Associate Professor April - June 2020 Curtin University of Technology Sarawak, Malaysia FOUNDER EDITOR-IN-CHIEF Audra Ong Dr. Hari Dayal Gupta Professor of Accounting [email protected] University of Windsor Ontario, Canada EDITOR-IN-CHIEF Franco Gandolfi Brent M Wren Professor Professor of Marketing Georgetown University Associate Provost Washington DC, USA The University of Alabama in Huntsville, US Copyright © 2020 South Asia Publications E. Stephen Grant Professor of Marketing & Entrepreneurship Published by Associate Dean, Undergraduate Programs South Asia Publications Faculty of Business Administration University of New Brunswick Annual Subscription (India) Singer 255, Fredericton, Canada Online* INR 4720.00 Print** INR 5000.00 Erdogan Ekiz *Including Service tax Associate Professor **Print includes complimentary online limited access. Dean, School of Hospitality Business and Management Mohammed VI Polytechnic University Online available at Morocco www.ebsco.com www.indianjournals.com Halil Dincer Kaya www.managementresearch.co.in Professor of Finance Northeastern State University Available in following EBSCO products: Broken Arrow, OK, USA Advanced Placement Source Business Source Alumni Edition Jin-Li Hu Business Source Complete Institute of Business and Management Business Source Corporate National Chiao Tung University Business Source Corporate Plus Taipei City, TAIWAN Business Source Elite Business Source Main Edition Jitendra Mahakud Business Source Premier Associate Professor of Economics & Finance Business Source Ultimate Department of Humanities & Social Sciences IT Source Indian Institute of Technology, Kharagpur Leadership & Management Source West Bengal, India Leadership & Management Source Plus Military Transition Support Center Marcus L. Stephenson NSA Collection Professor of Hospitality Management The Belt and Road Initiative Reference Source Dean, School of Hospitality Sunway University, Malaysia Printed at: Script Makers, 19, A1-B, DDA Market, Varun Gupta Paschim Vihar, New Delhi 110063 Associate Professor of Operations Management University of Chicago Booth School of Business Chicago, IL, USA Editorial Office South Asia Publications 340 SFS Flats Ashok Vihar Phase - 4 Delhi - 110052, India

Journal of Management Research Vol. 20, No. 2, April - June 2020, pp. 63-74 Integrated Framework for Vendor Management: An Interpretivist Approach Sanjeev Verma and Prateek Girdhar Abstract Studies show that many Indian construction projects are facing time and cost overrun. Vendor management is crucial for timely and within cost completion of any construction project. We used the interpretivist approach in this study to discover the critical parameters for effective vendor management. The grounded theory method with content analysis and thematic analysis helped in the development of the integrated framework. The integrated conceptual framework includes antecedents, moderators, and mediators of vendor management. The findings of the study are expected to play an instrumental role in improving the project management paradigm. Keywords: V endor Management, Strategic Management, Qualitative Research, Grounded Theory, Content Analysis, Thematic Analysis INTRODUCTION stakeholder circle. Stakeholder theory include “In any organization, a stakeholder is a group or three aspects of vendor management viz. normative an individual that is affected by the achievements validity, contributory power, and descriptive of the organization’s objectives or are in a position accuracy (Donaldson & Preston, 1995). Although to affect them” (Freeman, 1984, p. 46). Few three aspects of stakeholder theory are different stakeholders, like employees and customers, are perspectives of vendor management, they all emerge crucial for commercial existence as they offer the from the same basic premise. Donaldson & Preston organization with essential resources (Pfeffer & (1995) suggested that all three facets of stakeholder Salancik, 1978). Besides offering essential resources, theory are equally essential and complementary to some stakeholders, like vendors of essential supplies, each other. The core philosophy of stakeholder have the power to influence the project progress and management entails holistic project and value chain project outcome. Thus, it is essential to use holistic management. vendor management for the successful execution of The value chain is the backbone of project the project. management and, if properly managed, can Multiple studies attempted to study stakeholder yield superior project performance. The value management in the existing literature and covered chain includes multiple stakeholders ranging a broad range of topics like stakeholder theory, from bidders, sponsors, owners, vendors, project stakeholder analysis, stakeholder culture and managers, project teams, and customers. On the upstream side, vendors have the power to influence Sanjeev Verma the project, both positively and negatively. So, Dept of Marketing , NITIE vendor assessment is a critical part of vendor Vihar Lake Road, Mumbai-400087, India management. Vendor assessment is a system of Prateek Girdhar ranking and recording the performance of a vendor Dept of Marketing , NITIE on diverse issues that might include quality, delivery Vihar Lake Road, Mumbai-400087, India performance, price competitiveness, ease of doing

business, and supplier capability. The effectiveness The stakeholders can be primary or secondary, of the purchasing department is judged by the depending on their influencing power (Clarkson, quality and reliability of its vendors. 1995). The primary stakeholders can affect the Vendors are important stakeholder of an project directly and have a high level of interest in organization (Donaldson & Preston, 1995), and the project. They have the power to influence the vendor management is critical for the construction decisions in favor of the project and are given vital industry as any lapse in it leads to time and cost importance (Kanter, 1999). The primary stakeholders overruns. A systematic review of previous studies include the project manager, project management on vendor relationship management revealed team, other project team members, and project that most of the existing literature focused on sponsors (Galbreath, 2006). On the other hand, the positivist view of vendor management. The the secondary stakeholders have a lesser interest in positivist view uses a quantified approach for the project compared to primary stakeholders, but understanding vendor management and leaves they can also influence the project significantly. The the qualitative aspects. Relational dimensions are secondary stakeholders are functional managers, more qualitative, and quantification of relational customers, financial institutions, operations dimensions fails to uncover deep psychological teams, consultants, portfolio managers, and other aspects of vendor management. Interpretivism is an stakeholders (Donaldson & Preston, 1995). alternative approach to examine these psychological The success and failure of any project depend on aspects that have remained unexplored in the stakeholders’ engagement in value creation by vendor management literature. To fill this research the project (Bourne, 2005). To identify the key gap, we have used the interpretivism approach in stakeholder’s engagement during various stages this study for the development of an integrated of the project’s lifecycle, Weaver and Bourne vendor management framework. The findings of (2002) propounded the stakeholder circle concept. this study will help the construction project firms Stakeholder’s circle helps in understanding the to understand the crucial parameters of vendor expectations, engagement strategy, and relative management. influence of various project stakeholders. Stakeholder’s impact depends on proximity, LITERATURE REVIEW urgency, and power of a stakeholder (Cleland Stakeholders are interested entities as well as the & Ireland, 1999). Stakeholder’s circle includes members of the project team, external or internal concentric circular lines that depict the proximity of to the organization. The project team classifies the stakeholders and relative influence. The radial depth stakeholders into various categories, like internal gauges the degree of impact, colors, and patterns that or external, to regulate the expectations of all the indicate the outcome of the project (Bourne, 2005). involved parties (Donaldson & Preston, 1995). To Stakeholder Theory ensure effective vendor management, the project The stakeholder model classified the stakeholders manager maps the role and significance of these into various groups like investor, government, trade stakeholders on project performance. The success of associations, employees, political groups, suppliers, any project depends on the value added by a mutual customers, and communities (Donaldson & Preston, association between stakeholders and the project 1995). The stakeholder model describes the reasons team. Project performance is directly proportional why the stakeholder approach is essential to business to vendor management. The synergistic association and society (Bucholtz & Caroll, 2014). Three values between the project firm and its stakeholders of the stakeholder model are descriptive value, ensures the successful management of the project. instrumental value, and normative value. It depends on the ability of the project manager Three values of the stakeholder model, when how he manages the stakeholders and the dynamics arranged in the form of concentric circles, represent involved in the project. Journal of Management Research 64

the three facets of stakeholder theory viz. descriptive, can be plotted on the matrix. Table 1 classifies instrumental, and normative (Donaldson & Preston, the stakeholders based on power, legitimacy, and 1995). The outermost shell represents descriptive urgency. According to stakeholder classification values, the middle shell represents instrumental and identification scheme, the stakeholders are value, and the innermost shell represents normative considered as high priority stakeholders if they values. According to Donaldson and Preston possess the above three attributes, low priority (1995), a descriptive shell explains corporations stakeholders if they possess two attributes and least as a constellation of co-operative and competitive priority ones if they possess one attribute. Non- interests possessing intrinsic values, Stakeholder stakeholders possess none of the three attributes model can play an instrumental role in the (Jawahar & McLaughlin, 2001). achievement of business goals (profitability, growth, and sustainability). Stakeholders have legitimate Stakeholder Approach stakes that incorporate activities based on their Clear communication between the firm and its intrinsic interests and values. Normative values stakeholders sets the tone for effective project encapsulate the incorporation of legitimate stakes management. Communication strategy with of stakeholders (Donaldson & Preston, 1995). The proper agenda, deliverables, and targeted audience descriptive aspect explains the relation noticed in the minimizes chaos and disorder. According to Darter external world and is supported by instrumental and (2014), stakeholders can be approached with three normative value at the inner level. The presumption paths: least resistance, shotgun approach, and take that stakeholder management leads to better the high road path. Brief description of each path to economic performance is incomplete. At the core, approach stakeholders is as follows acknowledgment of moral values and obligations a) Least Resistance – This path is the most natural gives stakeholder management its central normative base. path taken by the project manager. He decides to inform his immediate superior or the critical Stakeholder Classification and Identification stakeholder and let them take the information According to Frooman (1999), four classes of to the other stakeholders. stakeholders influence project management b) Shotgun Approach – This approach is not limited strategies. Based on the resource dependency theory, to sharing of information with only one person the authors investigated four types of strategies or key stakeholder, but it is used to deliver the viz. direct withholding stakeholders, indirect message to more stakeholders in one go like withholding, direct usage, and indirect usage. Based corridor conversations, speaking out loud, etc. on the kinds of resource relationships, stakeholders Table 1: Classification of Stakeholders Types of Stakeholders Classification Scheme Latent stakeholders (possessing Dormant stakeholder: Possess power and lacks legitimacy and urgency single attribute) Discretionary stakeholder: Possess legitimacy and lacks power and urgency Demanding stakeholder: Possess urgency and lacks power and legitimacy Expectant stakeholders (possessing Dominant stakeholder: Possess power and legitimacy but lacks urgency two attributes) Dangerous stakeholder: Possess power and urgency but lacks legitimacy Dependent stakeholder: Possess legitimacy and urgency but lacks power Definitive stakeholder (possessing Possess all three attributes, i.e., power, legitimacy, and urgency three attributes) Non-stakeholder Possess none of three attributes, i.e., power, legitimacy, and urgency Volume 20, Number 2 • April - June 2020 65

c) Take the High Road – This method lies between have low power to influence the project, but the above two methods. This is the more formal their scope of participation is high. process that aims to conduct formal meetings c) 12 % of the stakeholders are contributing to with the stakeholders. collaboration issues for a project, i.e., these stakeholders have high power to influence the Corporate Governance project, but their scope of participation is low. Corporate governance is a system that directs & d) 61 % of the stakeholders are contributing controls the functioning of organizations (Cadbury, towards dialogue & issues advisory for a 2000). Conventionally agency concept governed the project, i.e., these stakeholders have low power aim of corporate governance (Jensen & Meckling, to influence the project, and their scope of 1976; Williamson, 1975) as the development of participation is also low. returns for shareholders (Friedman, 1970). From the political viewpoint (Scherer & Palazzo, 2007), Relationship Marketing and Role of corporate governance depends on the consensus Stakeholders given by the governed. It refers to the independent Recently with the expansion of global markets, values such as the separation of powers and political a surge has been seen in relationship marketing debate (Gomez et al., 2005). based strategic approaches based on the use of The power grid of stakeholders classifies the power new communication technologies, information, and interest of stakeholders in the form of a matrix and ongoing deregulation of many industries. varying from low to high (Spitzeck & Hansen, Previous studies have identified numerous types 2010). The X-axis presents the level of interest of industrial and business relationships. According in the project while the y-axis presents the level to Donaldson and Vallone (2002), understanding of power. If the stakeholder has a high interest in of value proposition is required for fostering a the project and has a high power to influence the better relationship. The relationship marketing project, then it is classified as the key stakeholder. literature acknowledges the importance of long- If the extent of the interest is low, but the ability to term relationships between customers and influence the project is high, then such stakeholders’ shareholders group. Recent research has also must be kept satisfied. If the scope of interest is emphasized the importance of the relationship high and the power to influence the project is low, between an organization and its associated then the stakeholders must be kept informed. In partners (Hennig-Thurau et al., 2000; Sheth & case interest is low and the level of power is also Parvatiyar, 1995). Stakeholder relationship has low, then minimum effort must be taken to satisfy gained significant attention in recent years due to the stakeholders. its overarching effect on the overall performance of According to Spitzeck and Hansen (2010), an organization (Christopher et al., 1991; Doyle, stakeholder governance is classified in the ratio of 1995; Gummesson, 1995; Morgan & Hunt, 1994; their participation in the decision making and made Buttle, 1999). the following observations To place strategic importance on different a) 14 % of the stakeholders are contributing stakeholder’s relationships, the “six markets” model classifies stakeholders in six market domains (Payne towards strategic collaboration for a project, & Holt, 2001). According to Christopher et al. i.e., these stakeholders have high power to (1991), customer markets are classified into the influence the project, and their scope of following six types: participation is also high. a) Customer Markets (includes prospective and b) 13 % of the stakeholders are contributing towards strategic advisory and innovation existing customers as well as intermediaries) decisions for a project, i.e., these stakeholders 66 Journal of Management Research

b) Referral Markets (these comprises two main of participants varied from graduation to post- categories – referral sources, such as existing graduation and had occupations ranging from an customers who refer their suppliers to others advisor, procurement managers, project managers, and multipliers that refer work to some outer and other significant people from middle and firm) top management. We intimated the participants beforehand for the interview meeting. The personal c) Influencer Markets (they include shareholders, interview lasted from 45 minutes to an hour. With financial analysts, government, business the prior consent of the participants, interviews process, and the consumer groups) were recorded for further analysis. A structured scheme of relevant questions was prepared for open- d) Employee Markets (the main concern is to ended interviews, and the laddering up technique attract the right employee to the organization) was used for gauging the real intent to manage vendors effectively. e) Internal Markets (includes staff and internal Data was transcribed and processed with the department including the organization) help of NVIVO software. For data analysis, word frequency count and word cloud technique were f ) Supplier/Alliance Markets (includes traditional used to identify the most critical factors. Further, suppliers and organizations with strategic for content analysis, data were analyzed using alliance) primary coding, secondary coding, and tertiary coding. Constant comparison method was used for RESEARCH DESIGN the identification of emergent themes. Intercoder This study attempts to identify crucial factors for reliability and other norms of validity in qualitative vendor management in the Indian Construction research were maintained. Industry. These factors are playing an interdependent role as an antecedent, mediator, RESULTS AND DISCUSSION and moderator for subsequent vendor management. Initially, data was analyzed and used for descriptive Exploratory study research design based on primary analysis. Figure 1 shows the distribution of data is used for building the conceptual framework. respondents based on their designations. Multiple positions like procurement manager, project Open-ended, in-depth interviews, along with managers, engineer, consultant, project officer, etc. thematic content analysis, were used to derive the are involved in dealing with vendors for various underline constructs and their interrelationship. purposes. To capture the views of all key positions, Content analysis and thematic analysis were used the respondents in this study range from engineers for the identification of variables, and the grounded to managers. Out of 28 respondents, 17 were theory method was used for building the proposed managers, 7 were engineers, 3 were consultant, and conceptual framework. 1 respondent was project officer. 17 respondents in the manager category had different managerial The research study adopted the qualitative research responsibilities: 6 were general managers, 4 were design focusing on data collected from a sample of project managers, 4 were procurement managers, 28 construction companies established after 1922 and 3 were the assistant managers. The kind and in India. Participants, each with experience of level of interaction between managers and vendors more than 36 months, were purposively selected vary based on coordinating manager specialization. from the respective construction companies and Procurement managers and project managers have interviewed. Respondents include 5 females and more detailed and focused communication with 23 males with their ages ranging between 28 years vendors. Out of 7 engineers, 4 were senior engineers, to 52 years. The rationale behind selecting the participants was that they possessed the relevant experience useful for identifying the parameters crucial for vendor management in the Indian construction industry. The education qualification Volume 20, Number 2 • April - June 2020 67

and 3 were project engineers. The interaction respondents had 59 to 79 months experience, 6 between engineers and vendors was more technical. respondents had 79 to 139 months of experience, The specification of materials, dimensions and and only 1 respondent had more than 139 months quality were discussed during the technical discussion of experience. It clearly shows that most of the between engineers and vendors. Terms of the time, young and lesser experienced personnel contract and delivery period etc. were discussed more interact with the vendor. The majority of the often with the managers. Consultants and advisors respondents (70 percent) had less than 79 months bring the external competencies for smooth sailing (12.5 years) of experience who were entrusted to of projects, and their rich experience helps in the manage vendors. It’s a fascinating phenomenon identification of the right vendor for the specific job. to observe, and it points at proficiency possessed Further, respondents were classified based on their by young and lesser experienced people to manage experience. The experienced personnel had a more vendors. in-depth understanding of vendors’ expectation Drawing upon the NVivo 11 results, Figure and various aspects of vendor management. Due 3 presents the cluster analysis of the top 30 to their more extended stay with the construction parameters that emerged for effective vendor organization or project, they were able to develop management. Some of the critical emergent mutually beneficial relationships with the vendors. parameters for effective project management are The varying views of respondents based on delivery time, quality, cost, location, previous their experience provided an array of ideas for experience, and capacity to supply. better vendor management. Figure 2 presents the distribution of respondents based on their Antecedents of Vendor Management experience expressed in the number of months. The participants were asked to identify the factors Out of 28 respondents, 23 respondents shared which, according to them, are crucial for vendor their experience duration while 5 respondents management in the Indian construction industry. hesitated to share their experience duration. Each participant gave inputs based on their past 10 respondents had 59 months of experience, 6 experiences in the industry. The qualitative data Number of Participants 7 6 5 4 3 2 1 0 Designation FFigiguurree: 11:DDisisttrriibbuuttiioonn ooff PPaarrttiicciippaanntstsbBaaseseddoonnDDeesisgignnaatitoinon 68 Journal of Management Research

obtained was screened, cleaned, and tabulated for the open codes, axial codes, and thematic codes. further analysis with the help of word frequency Table 2 shows the complete list of parameters count. The narrative analysis was done to identify identified as antecedents by the participants. Number of Participants Experience in Months FiguFrigeu2re: 2R:eRspesopnonddeennttss’EExxppeerireinecnec(eIn(IMnoMntohns)ths) Figure 3: Cluster Analysis of Top 30 Parameters 69 Volume 20, Number 2 • April - June 2020

As evident from Table 2, previous experience of the construction project requirement and align the vendor and pricing are the most significant seamlessly with the project management team. antecedent for effective vendor management. Project Pricing is the second most important dimension management executives emphasize the importance that emerges for effective vendor management. of the selection of the right vendor for a given job Pricing here does not focus on a lower price and mention the significance of previous experience (monetary value), but the price here indicates value for smooth working. Experienced vendors know Table 2: List of Antecedents Open Codes Word Cumulative Axial Code Freq. Freq. Payment Terms Annual Turnover 2 10 Pricing Credit Terms 1 Financials 1 10 Experience and Expertise Price 1 Negotiations 3 7 Capability Profit Margin 1 Previous Experience 1 5 Quality Knowledge 6 4 Terms and Conditions Performance 1 4 Time References 1 4 Brand/Market Identity Technical Expertise 1 2 Place Capability 1 Handling 1 Government Support 1 Organization Structure 1 Crisis Management 1 Infrastructure 1 Database Management 1 Quality 1 Quality Management 2 Quality of Product 1 Rating 1 Service Agreement 1 Review Meeting 1 Norms 1 Order Acknowledgement 1 Delivery Time 1 Schedule Preparation 2 Timely Delivery 1 Market Trend 1 Reputation 1 Approach 1 Reliability 1 Location 1 Availability 1 1 70 Journal of Management Research

concept of price. The better financial standing of and communication means, place factor is losing the vendor provides an opportunity for negotiation, its importance, and vendors may compete for all credit limit, and relaxation in payment terms. national and international projects. The vendor with deep pockets maintains the consistency and quality for the building of long- Mediators for Vendor Management term relationships with construction companies as The definition of the mediator was briefed to they know one successful project opens the avenue respondents before asking the question related for multiple new project collaborations. to the mediator. Mediator was defined as an Capabilities are the third most important intervening variable between an independent and determinant of effective vendor management dependent variable. For clarity purposes, examples after the pricing and experience of the vendor. of mediators were given in different contexts so that Infrastructural capabilities, information technology respondents can think of mediating factors affecting capabilities, crisis management, and dynamic vendor management in the construction project flexibility are some of the much-emphasized organization. characteristics appreciated by construction project When the participants were asked to identify the organizations. mediators, which, according to them, are crucial Quality, terms and conditions, timeline, and for vendor management in the Indian construction market identity of vendors are some of the other industry, the emergent mediating factors mentioned vital aspects closely scrutinized by project personnel. are presented in Table 3. Broadly, terms and conditions include both quality Emergent open codes were classified into axial parameters and time frame for the delivery of codes, and significant mediating variables can the job. Interestingly, the market identity of be grouped into 3 categories. The most often vendors is also becoming a prominent factor in mentioned mediating factors were the site team. vendor management. The referrals from similar Respondents feel that team composition responsible construction companies play a crucial role in the for project execution, team passion, and site support identification of the vendor for the job at hand. In plays a significant mediating role in managing one of the narratives, the respondent pointed at a vendors. The passionate team induces enthusiasm bid by invitation to the high-rank vendors only. in vendors and treats vendors as an important team The geographical location (place) is not significant member for the project’s success. in vendor selection and management process. The second mediating variables indicate the Construction project organizations wanted to have importance of project charter and KPI in developing the best vendor irrespective of their location, and a productive relationship with vendors. Educating regional influences may not deter their choice for vendors about project charter and KPI improves the the best possible vendor. With better transportation Table 3: List of Mediators Open Code Word Cumulative Axial Code KPI Freq. Freq. Project Charter Passion 5 8 Project Goal Site Support 3 Team 4 10 Project Team 3 3 Volume 20, Number 2 • April - June 2020 71

vendor’s latitude to perform better and improves the paradigm uses a quantitative approach to feeling of inclusion. Higher the imbibition of the understand the phenomenon, but the deep-rooted project charter and KPI among vendors, higher is the latent need of vendors remains unexplored. This chances of effective vendor management. study used the interpretive approach to discover the potential factors for effective vendor management Moderators for Vendor Management in ever-changing time. The fast-changing needs Moderator variables affect the strength of the and aspirations of vendors bring contemporary relationship between independent variables and challenges, and relationships are becoming more dependent variables. When we asked participants fragile. As cited by previous research, vendors are to identify the significant moderators for one of the most critical stakeholders in the success vendor management, respondents indicated the of any project, and effective vendor management importance of grievance redressal in effective project is a must for effective project management. This management. Table 4 presents the open codes exploratory study brings insights that can be representing the moderator variable. instrumental in reshaping the project management Narrative analysis hinted at the apparent emergence strategies for improved results. of the difference of opinion between vendors and We used the qualitative approach with an in- construction project organization, but regular depth interview and ethnography in this study feedback and prompt action can resolve grievance for data collection. The grounded theory without creating any scuffle between the two. Both method with content and thematic analysis antecedents and mediating variables have indicated helped in the development of an integrated the integrated role of vendors in successful project conceptual framework. Figure 4 presents the completion. So, effective grievance resolution can integrated framework depicting the relationship moderate the strength of the relationship between between antecedent variable, mediating variable, vendors and project organization. moderating variable, and effective vendor management. Pricing and previous experience CONCLUSION of vendor emerged as the most crucial criterion Vendor management is always a challenging area variable for effective vendor management. The of concern for most of the project management value proposition through appropriate pricing organizations. Previous studies have provided and expertise of the vendor can make them the battery of variables having a significant indispensable for project organization. The role in building and maintaining an effective profit-making should not be aim of the vendors. relationship with vendors, but out of box thinking Instead, building a reputation for long term and understanding latent need were unexplored. relationship helps the vendor to garner positive Existing literature review revealed that past studies referral and makes them an integral part of overall used the positivist paradigm for understanding project management. Construction companies vendor management phenomenon. The Positivist undertaking a significant civil project can reap this benefit of vendor management by outsourcing many jobs to reliable vendors. Table 4: List of Moderators Open Code Word Cumulative Axial Code Dispute Resolving Freq. Freq. Prompt Action Feedback 6 14 Grievance Redressal 4 4 72 Journal of Management Research

Pricing Grievance Redressal Vendor Management Experience and Expertise Project Goal Capability Project Team Quality Capability Terms and Conditions Timeframe Market Identity Place Grievance Redressal Figure 4: Integrated Conceptual Framework Figure 4: Integrated Conceptual Framework The project team plays a significant mediating but magnificence lies in reducing the difference role between criterion variables and effective in opinion. The regular feedback mechanism and vendor management. The project team passion can induce vendors for higher performance goals and prompt action instill a sense of confidence in inculcate a sense of belongingness in the vendors. vendors. Construction companies should develop The feeling of unification for common project goal an effective grievance handling mechanism for between vendor and project team align efforts to a better vendor management. The lesser conflict leads more significant impact. The project management to smooth communication and lesser confusion, team may use management by objective (MBO) which in turn improves productivity. technique to bring more proximity between This study used the qualitative approach using vendors and the organization. The consultative and an in-depth survey. For more precise findings complementary approach provides an opportunity and triangulation of data, future researchers may for finishing the project within time and approved use other qualitative techniques like focus group budget. studies. Future researchers may also attempt to test Grievance redressal emerged as a significant the proposed conceptual framework with the help moderator between vendor and project organization. of quantitative methods. For a broader application, The difference of opinion is noticeable when a the integrated framework may be tested in different human organization works with a different culture, kinds of project organizations. REFERENCES Bourne, L., & Walker, D.H.T. (2005). Visualizing and mapping stakeholder influence. Management Decision, 43, 649-60. Buttle, F. (1999). The SCOPE of customer relationship management. International Journal of Customer Relationship Management, 1(4), 327-337. Volume 20, Number 2 • April - June 2020 73

Cadbury, S. A. (2000). The corporate governance agenda. Corporate Governance: An International Review, 8(1), 7-15. Carroll, A., and Bucholtz, A. (2014). Business and Society: Ethics, Sustainability, and Stakeholder Management. South-Western College Publishing Christopher, M., Payne, A., & Ballantyne, D. (1991). Relationship marketing: bringing quality customer service and marketing together. Clarkson, M.B.E. (1995). A stakeholder framework for analyzing and evaluating corporate social performance. Academy of Management Review, 20, 92-117. Cleland, D. I., & Ireland, L. R. (1999). Project management: strategic design and implementation (Vol. 4). Singapore: McGraw-Hill. Donaldson, S. I., & Grant-Vallone, E. J. (2002). Understanding self-report bias in organizational behavior research. Journal of Business and Psychology, 17(2), 245-260. Donaldson, T., & Preston, L. E. (1995). The stakeholder theory of the corporation: Concepts, evidence, and implications. Academy of Management Review, 20(1), 65-91. Darter, K. (2014). Three paths to approaching stakeholders. Retrieved from https://www.projectsmart.co.uk/three-paths-to-approaching- stakeholders.php on April 23, 2020. Doyle, P. (1995). Marketing in the new millennium. European Journal of Marketing, 29(13), 23-41. Freeman, E. (1984). Strategic Management: A Stakeholder Approach, Pitman, Boston, MA. Friedman, M. (1970). A theoretical framework for monetary analysis. Journal of Political Economy, 78(2), 193-238. Frooman, J. (1999). Stakeholder influence strategies. Academy of management review, 24(2), 191-205. Galbreath, J. (2006). Corporate social responsibility strategy: strategic options, global considerations.  Corporate Governance: The international journal of business in society, 6(2), 175-187. Gomez, P., Céspedes-Lorente, J., & Valle-Cabrera, R. (2005). Organizational learning capability: a proposal for measurement. Journal of business research, 58(6), 715-725. Gummesson, E. (1995). Relationship marketing. Gower Handbook of Marketing, 4. Hennig-Thurau, T., & Hansen, U. (2000). Relationship marketing—Some reflections on the state-of-the-art of the relational concept. Relationship marketing (pp. 3-27). Springer, Berlin, Heidelberg. Jawahar, I. M., & McLaughlin, G. L. (2001). Toward a descriptive stakeholder theory: An organizational life cycle approach. Academy of management review, 26(3), 397-414. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs, and ownership structure. Journal of financial economics, 3(4), 305-360. Kanter, R. M. (1999). From spare change to real change. The social sector as a beta site for business innovation. Harvard business review, 77(3), 122-32. Morgan, R. M., & Hunt, S. D. (1994). The commitment-trust theory of relationship marketing. The journal of marketing, 20-38. Payne, A., & Holt, S. (2001). Diagnosing customer value: integrating the value process and relationship marketing. British Journal of Management, 12(2), 159-182. Pfeffer, J., & Salancik, G. R. (1978). The external control of organizations: A resource dependence approach. NY: Harper and Row Publishers. Scherer, A. G., & Palazzo, G. (2007). Toward a political conception of corporate responsibility: Business and society seen from a Habermasian perspective. Academy of management review, 32(4), 1096-1120. Sheth, J. N., & Parvatiyar, A. (1995). Relationship marketing in consumer markets: antecedents and consequences. Journal of the Academy of Marketing Science, 23(4), 255-271. Spitzeck, H., & Hansen, E. G. (2010). Stakeholder governance: how stakeholders influence corporate decision making. Corporate Governance: The international journal of business in society, 10(4), 378-391. Weaver, P., & Bourne, L. (2002, October). Projects-Fact or Fiction. In PMI Conference-Maximizing Project Value, Melbourne, PMI Melbourne Chapter. Williamson, O. E. (1975). Markets and Hierarchies: Analysis and Antitrust Implications: A Study in the Economics of Internal Organization. University of Illinois at Urbana-Champaign’s Academy for Entrepreneurial Leadership Historical Research Reference in Entrepreneurship. 74 Journal of Management Research

Journal of Management Research Vol. 20, No. 2, April - June 2020, pp. 75-98 Explanation and Rigor in Management Theorizing: Which Theory-Building Criteria Make for an Influential Contribution? Jaume Franquesa, William Acar and Jino O. Mwaka Abstract This paper advances and tests a comprehensive but parsimonious model of theory-building/evaluation criteria in management and organization science, and of the relationships between these criteria and a theory’s eventual prominence within the discipline. The model is tested using survey data in which knowledgeable scholars are asked to provide a detailed assessment of the traits of one of a few well-known seminal theoretical articles that are used as vehicles. The results support the presence of three distinct but correlated dimensions of theory evaluation (novelty, extendibility, and relevance to practice) and further provide confirmatory evidence of an overarching, second-order construct, which we term the explanatory meaningfulness of a theoretical exposition. Moreover, we find this construct to be a positive and strong predictor of the subsequent perceived importance of a theoretical article among management scholars. By contrast, the logical consistency and the falsifiability of the theoretical exposition were not significantly associated with its perceived importance. Paradoxically, our findings suggest that the most influential theoretical articles in management are those that offer greater explanatory value (stemming from their originality and perceived usefulness for research and practice), regardless of other aspects associated with conventional prescriptions for rigorous theory-making. Keywords: M anagement Theory, Theory Building, Theory Evaluation, Explanatory Meaningfulness, Scientific Rigor, Scholarly Impact INTRODUCTION questions (Kaplan, 1964), and offering insightful predictions (Hambrick, 2007). Similarly, theories Theories are central to the advancement of assist in organizing and making sense of what knowledge in any field of scientific inquiry. would otherwise be disconnected propositions They guide the research agenda by organizing and empirical findings, thus facilitating the the complex empirical world and pointing to effective accumulation of knowledge (Haveman, the important questions (Bacharach, 1989), Mahoney, & Mannix, 2019; Turner, 1985). The suggesting appropriate experiments to answer such precise definition and desirable qualities of theory, however, are matters of some dispute within the Jaume Franquesa (Corresponding Author) organization studies and management literature. James F. Dicke College of Business Administration There have been long-lasting and recurrent debates Ohio Northern University, Ada, OH 45810 about what constitutes a theory (e.g., DiMaggio, William Acar 1995; Shapira, 2011; Sutton & Staw, 1995), Graduate School of Management and about the requisite elements of a “good” Kent State University, Kent, OH 44242 theory and thus the criteria for the evaluation Jino O. Mwaka of theoretical contributions (e.g., Ackoff, 1962; University of the Sacred Heart Bacharach, 1989; Corley & Gioia, 2011). Gulu, Uganda

In broad terms, the different schools of thought unquestionable value of these contributions, they in the theory building/theory evaluation literature suffer from two important limitations: First, in all in management can be organized as (i) those that cases, they are circumscribed to only a few (and focus on the explanatory value or usefulness of broadly defined) theory evaluation criteria under account of the theoretical exposition (e.g., Astley investigation at a time, so they offer rather partial & Zammuto, 1992; Corley & Gioia, 2011; Gioia and tentative evidence. Second, these studies have & Pitre, 1990; McKinley, Mone, & Moon, 1999; tended to measure the variables of interest as ex post Sutton & Staw, 1995); (ii) those that focus on the outcomes—that is, some years after the publication Popperian concept of falsifiability and/or other of the theoretical work in question—as opposed conditions for rigorous construction as well as to as characteristics of the theoretical treatise itself exacting empirical tests of theories (e.g., Arend, (e.g., in Miner, 1984, 2003). As such, they do not Sarooghi, & Burkemper, 2015; Dubin, 1969; Miller inform theorists and their evaluators of the desirable & Tsang, 2010; Shapira, 2011); and (iii) those that qualities during theory-making which will drive the place similar importance on both explanatory value ensuing prominence of a theoretical work. and scientific rigor/validity requirements (e.g., This paper makes several contributions to Acar, Franquesa, & Mwaka, 2020; Ackoff, 1962; this literature. First, we synthesize the extant Bacharach, 1989; Eisenhardt, 1989b; Whetten, conceptual literature to propose a comprehensive 1989). The criteria emphasized by these differing yet parsimonious model of ex ante theory-building/ perspectives are only partially overlapping and, most evaluation criteria that are likely to drive a theory’s importantly, often at odds with each other (i.e., interest, adherence and, ultimately, salience within there are trade-offs in the simultaneous pursuit of the management academic community. Second, divergent sets of criteria). we introduce a new, overarching theory-building/ Empirical studies may advance the debates by evaluation construct, which we term the explanatory shedding light on the criteria associated with lasting meaningfulness of a theoretical exposition, and and influential theories in management. Besides provide confirmatory evidence in support of this considerations like the reputation of the authors or higher-order latent variable. Third, we advance and the journal of publication (Cole & Cole, 1967; Judge validate a survey instrument that operationalizes et al., 2007), which characteristics of a theoretical key theory evaluation constructs of interest in the contribution are most strongly associated with its prior literature. Finally, we explore the relationship eventual prominence? Are scientific rigor/accuracy between the ex ante theory evaluation traits of aspects equally appreciated by management scholars theoretical contributions and the eventual perceived as aspects related to richness/usefulness of account, or scholarly importance of such contributions, thus does one type of consideration prevail over the other? pointing to recommendations for effective theory Beyond the personal viewpoints of various authors building by management scholars. (e.g., Dubin, 1969), as well as the evolving and The rest of the paper is organized as follows: We varying guidelines offered in editor notes of leading begin by summarizing the embryonic empirical journals (e.g., Barney, 2018; Kilduff, 2006; Whetten, literature on the relationships between satisfaction 1989), what empirically validated recommendations of theory evaluation criteria and eventual can we offer to management researchers for effective prominence of theoretical contributions among theory building? management scholars. Next, we draw from the To date, only a few studies have begun to explore prior conceptual literature to develop our proposed the relationships between qualities of a theoretical model of key theory-evaluation dimensions that contribution and its subsequent scholarly impact drive the subsequent importance of theoretical within the management field (Colquitt & Zapata- works. Ensuing sections describe, in turn, the Phelan, 2007; Miner, 1984, 2003). Despite the methods used and the results of the empirical test of our model. We conclude with a discussion of our 76 Journal of Management Research

findings, as well as of limitations of the study and the extent to which it adds to existing theory by implications for future research. proposing a new mediator or moderator, a new relationship, or even a new construct) and (ii) PRIOR EMPIRICAL RESEARCH AND theory testing (i.e., the extent to which predictions/ FINDINGS hypotheses are grounded on something closest to a “true theory”). The authors’ theory building and Three prior studies have explored relationships testing ratings were both found to be positively between theory evaluation correlates and the associated with an article’s scholarly impact, as subsequent scholarly prominence of theoretical measured by citation rates. contributions in management: Miner (1984) Colquitt and Zapata-Phelan also found that the analyzed 32 management theories and explored types of empirical articles that received the most if (i) scientific validity (defined as the extent to citations were those with substantial and balanced which subsequent empirical tests had been found theory building and testing ratings, where articles generally supportive; as measured by the author’s in the mid-range of both variables achieved as much own ratings) and/or (ii) practical usefulness (defined prominence as those with high scores in both. From as the extent to which useful applications to this, and based on the conceptual theory-evaluation management practice had resulted; also subjectively literature (e.g., McKinley et al., 1999), they rated by the author) were related to a theory’s concluded that a balance between novelty on the scholarly importance (as measured by frequency one hand and continuity with an existing theoretical of nomination by a panel of knowledgeable formulation on the other helps bring the most scholars). Surprisingly, neither scientific validity nor attention to an article among management scholars usefulness in practice appeared to be related to the (2007, p. 1293), regardless of whether such extant scholarly consensus regarding a theory’s importance. formulation was a vague framework, a model, or a In a re-test of the same relationships 20 years later, “true theory.” Miner (2003) broadened his study to a “reasonably To further explore predictors of scholarly complete listing” of 73 organizational behavior importance, our study builds upon the arguments theories. A panel of knowledgeable scholars provided and findings from the above investigations, as well importance ratings for each of them, while the as other conceptual theory evaluation work, albeit independent variables were measured in a similar using a novel empirical approach: First, while the fashion as in Miner (1984). This time, a significant data used in the above studies were single-rater correlation was found between the scientific validity assessments of a few traits on a broad swath of of a theory and its importance rating, while usefulness theories (Miner, 1984, 2003) or empirical articles in practice remained not significantly related to (Colquitt & Zapata-Phelan, 2007), our dataset importance. In light of this change in results, Miner is made of independent assessments by a large concluded that the field of organizational behavior sample of knowledgeable scholars of a larger set of had evolved into a “more mature science,” albeit traits from a few seminal (and hence memorable) one that might “have become too academic” at the theoretical exemplars which serve as vehicles. Our expense of disregarding “the matter of practical data were collected from hundreds of members of application” (2003, p. 262). pertinent divisions of the Academy of Management, Subsequently, Colquitt and Zapata-Phelan (2007) who were asked to complete a detailed assessment investigated the “theoretical contribution” made of just one or two theoretical papers with which by empirical articles in management by coding all they professed to be conversant. Second, while papers published in the Academy of Management measurement of explanatory variables in prior Journal between 1963 and 2007.1 The theoretical studies relied on single-item classifications derived contribution of an empirical article was defined from complex coding schemes, we advance and and coded as its levels of (i) theory building (i.e., validate survey-based measurement scales for the Volume 20, Number 2 • April - June 2020 77

key constructs of interest. Our study focuses on al., 2014; Conlon, 2002; Davis, 1971; McKinley uncovering the underlying relationships among key et al., 1999; Whetten, 1989). Thus, our construct traits of a theoretical treatise, as well as between encompasses two possible subdimensions of the latter and subsequent scholarly importance, as originality of a theoretical contribution, which opposed to seeking to develop comparative ratings Corley and Gioia (2011) respectively termed of different theoretical expositions. The next section “incremental” (i.e., progressive advances in develops our proposed conceptual model. an established course of understanding) and “revelatory” (i.e., surprising new perspectives or new PREDICTORS OF THEORY directions in understanding). IMPORTANCE Novelty may create defamiliarization, which DiMaggio (1995, p. 392) defined as the process It has been argued that the ability of a theoretical by which an academic discipline is led to see the contribution to explain important organizational world “with new eyes,” and which he argued phenomena is one of the determinants of its should be one of the aims of “good theory.” Yet, attractiveness, and ultimately salience, among it is important that novelty (even when it is of the management scholars (Astley & Zammuto, 1992; revelatory kind) not be overdone and that there McKinley et al., 1999). Based on our synthesis be some link or continuity with assumptions, of the theory building and evaluation literature rationales, constructs, and language already familiar in management (including commentary in the to scholars so that the theory can be understood editorial statements of leading journals), we argue and seen as legitimate (Davis, 1971; DiMaggio, that there are three distinct and fundamental 1995; McKinley at al., 1999). Thus, a certain indicators of the perceived explanatory value of a balance between defamiliarization and familiarity theory to management scholars: (1) its novelty (e.g., becomes an important consideration when Colquitt & Zapata-Phelan, 2007; Corley & Gioia, introducing novel theoretical aspects. 2011; McKinley et al., 1999), (2) its extendibility It should also be noted that, per the above or potential to generate further research (e.g., definition, our construct of novelty excludes McKinley et al., 1999; Whetten, 1989), and (3) theoretical invention that emanates from a lack of its relevance to practice (e.g., Corley & Gioia, 2011; appreciation of constructs and rationales already Miner, 1984, 2003). Moreover, we contribute available in the prior literature. This faux novelty that, although distinct, the above traits interplay contributes to a proliferation of redundant with one another in complex ways, so that the theoretical elements in our field, which muddies overall perceived value of a theory’s insights will the pursuit of understanding and undermines the be best captured as a unifying dimension. Thus, development of a coherent body of knowledge (e.g., we will propose below a new construct which we Oxley, Rivkin, & Ryall, 2010). term the explanatory meaningfulness of a theory Novelty, as defined here, could entail exploring (or of a model, framework, or other theoretical previously unexamined relationships among existing contribution) and which captures its overall a priori constructs, proposing new causal mechanisms informational value. for already known relationships, or modifying relationships within extant theories by adding new Novelty mediators or moderators. Novelty could also go Building on prior work, we define the novelty of further by exploring entirely new research questions a theory or theoretical exposition as the extent to involving novel phenomena and/or introducing which it challenges scholars’ extant thinking on brand-new constructs and justifying relationships organizational phenomena by either modifying among them (Bartunek et al., 2006; Colquitt & or extending current theories, or by offering an entirely new point of view (Barney, 2018; Bettis et 78 Journal of Management Research

Zapatta-Phelan, 2007; Davis, 1971; McKinley at as the range of phenomena to which the theory al., 1999). could be applied (Bacharach, 1989; McKinley et We expect novelty to be a driver of a theory’s al., 1999), as well as the possibility that it can be attractiveness and thus of its eventual prominence applied to different levels of analysis (Makadok among scholars. By separating its insights from the et al., 2018). As scope increases generalizability, multitude of frameworks, constructs, and variables it adds to theoretical relevance. It also increases that compete for a scholar’s attention, novelty helps the variety of operationalizations and measures the theory to get noticed (Colquitt & Zapatta- that may fit within the theory’s frame of reference Phelan, 2007; DiMaggio, 1995; McKinley at and, hence, the number of empirical studies that al., 1999). Novelty also raises the interest among could be presented as tests of the theory (Astley & scholars by creating opportunities for follow-up Zammuto, 1992; McKinley et al., 1999). Scope scholarship, which promotes adoption (McKinley at is fostered by the level of abstraction of a theory’s al., 1999). Consistent with this, a survey of editorial formulation, as well as by the adequate broadness board members of the Academy of Management of construct definitions and of the boundary Journal revealed that the most frequently cited conditions under which the theory applies reasons why an article was found interesting (Bacharach, 1989; McKinley et al., 1999; Suddaby, included, among others, whether it created new 2010). For example, Makadok et al. (2018, p. theory or whether it challenged established theory 1539) noted that when a theory’s assumptions or and created an “aha” moment (Bartunek et al., boundary conditions are very general, opportunities 2006). As a result of its effects on salience and exist to extend the theory to different sub-cases (i.e., interest, the novelty of a theory will contribute by narrowing conditions), in order to derive more to the eventual impact of the ideas it presents specific implications.2 Similarly, Bacharach (1989, (Whetten, 1989). p. 507) remarked that constructs and variables with broader scope increase the “explanatory power” of Extendibility the theory. The extendibility of a theory refers to the degree to A related, albeit less desirable, contributor which it provides opportunities for, and stimulates, of extendibility is the ambiguity of construct further theoretical development and/or broadened definitions, causal mechanisms, or boundary applicability (Acar et al., 2020). This includes conditions (Astley & Zammuto, 1992; McKinley opportunities to clarify or redefine its constructs, et al., 1999; Weik, 1995). We define a construct’s causal mechanisms, and/or boundary conditions, ambiguity as the extent to which the interpretation as well as opportunities to apply the theory to of its definition varies from scholar to scholar different phenomena or levels of analysis (Makadok, (Oxley et al., 2010). Likewise, ambiguity of Burton, & Barney, 2018). A good theoretical the causal mechanism refers to vagueness with contribution should offer an inspiring framework regard to the underlying logic that drives the that triggers subsequent development and energizes theory’s propositions. And ambiguity of boundary the exploration of new questions in ways that foster conditions refers to vagueness with regard to the the growth of knowledge (Bartunek et al., 2006; contexts and levels of analysis where the theory McKinley et al., 1999; Whetten, 1989). Thus, applies. Such ambiguities may be unavoidable in albeit different constructs, extendibility and novelty the initial stages of a new theoretical stream, but are mutually supportive: Novelty (especially of the they are also recognized as pervasive problems that revelatory kind) makes theories more extendible, plague management theory-making, in particular and extendibility breeds further novelty. due to our over-reliance on informal (Makadok Besides novelty, an important contributor to et al., 2018) or verbal (Oxley et al., 2010) extendibility is a broad ideational scope, defined modes of theorizing.3 Although unhelpful to the accuracy of knowledge claims, ambiguity provides Volume 20, Number 2 • April - June 2020 opportunities for refinements of the theory and, 79

thus, extendibility (McKinley et al., 1999; Suddaby, Drawing on Astley and Zammuto (1992), we define 2010). For example, Makadok et al. (2018, p. the a priori relevance to practice of a theoretical 1537) observed that ambiguity about a theory’s contribution as its potential to influence managerial boundary conditions offers an opportunity for (or other practitioners’) action through either subsequent research to debate and clarify its hidden (i) feasible tools and techniques that may derive assumptions. Similarly, other ambiguities may from the theory and will be directly applicable in provide opportunities to clarify a theory’s constructs practice or, more generally, (ii) concepts, ideas, and or underlying rationale, or to expose internal language that may increase practitioners’ ability inconsistencies in the latter (Makadok et al., 2018). to frame, analyze, and solve business problems, An additional contributor of extendibility is the as well as to legitimate chosen courses of action. richness of the theory, defined as the degree to Relevance to practice is also enhanced by (iii) the which it attempts to encompass different theoretical timeliness and importance to management practice perspectives and/or simultaneous organizational of the phenomena that the theory explains (Corley demands, even when inconsistent with one another, in & Gioia, 2011; Mintzberg, 2005; Ployhart & order to fully account for multifaceted organizational Bartunek, 2019). realities (DiMaggio, 1995; Gioia & Pitre, 1990; Lado Thus, our construct is broader than Miner’s et al., 2006; Poole & Van de Ven, 1989).4 “usefulness in practice” (1984, 2003), which focused Extendibility increases the utilitarian value of a only on practical tools or “applications” generated theory to scholars by facilitating and promoting a form the theory. Consistent with prior observations stream of theoretical and empirical research work, that organization science theories tend to be rather and publication, based on it. Indeed, we believe abstract and rarely reducible to specific managerial extendibility to be the key driver of the perceived tools (e.g., Beyer & Trice, 1982), Miner (2003, p. research usefulness of a theory. Thus, we expect 267) reported that most of the 73 theories in his extendibility to increase the scholarly appeal of study rated on the low end of his usefulness-in- a theory, leading to greater use, and ultimately practice scale. Indeed, the bulk of the contributions to heightened impact on the field in the form to practice by management theories may rather occur of a substantial cumulative literature (Astley & through the provision of conceptual language that Zammuto, 1992; Bacharach, 1989; Makadok et al., enhances the problem-framing and problem-solving 2018; McKinley et al., 1999). capabilities of managers and consultants (Astley & Zammuto, 1992).5 Relevance to Practice Based on our review of the theory building/ As a professional field, there is a long tradition evaluation literature, we expect that the relevance that organization and management theory should to practice of a theoretical contribution will have implications for, and effects on, practice increase its appeal among scholars, thus leading (Corley & Gioia, 2011; Hambrick, 1994, 2007; to its greater adoption and eventual impact on the Miner, 1984, 2003; Mintzberg, 2005; Sandberg field (Bartunek et al., 2006; Miner, 1984; 2003). & Tsoukas, 2011). Generating knowledge that is Consistent with this expectation, for example, usable “in the real world” and addressing subjects Corley and Gioia (2011, p. 19) reported that an that are timely and relevant to practitioners are seen observable pattern distinguishing the papers that as important factors that contribute to the quality received the Academy of Management Review Best of our theorizing (e.g., Kilduff, 2006; Ployhart & Article Award in any given year from the papers Bartunek, 2019; Whetten, 1989). As such, it is that ended up being the most cited over time (out often a requirement for publication in management of the same AMR volumes) is that the most cited journals that a theoretical contribution should be papers could also “be characterized as higher in “relevant to practice” (e.g., Rynes, 2005). utility for practice.” 80 Journal of Management Research

Overall Explanatory Value: The Concept of by a reflectively measured second-order factor. Explanatory Meaningfulness of a Theory/ Also, consistent with the arguments above for Model each of its component dimensions, we expect this Although distinct traits, we expect that the novelty, overarching second-order meaningfulness construct extendibility, and relevance to practice of a theoretical to be positively related to the eventual prominence contribution are connected elements that mutually of a theory among management scholars. Namely: reinforce each other. Given this, we propose Hypothesis 2: This tridimensional second-order that they will be best conceptualized as different construct, which we will refer to as the a priori manifestations of a single, tridimensional higher- explanatory meaningfulness of a theoretical treatise, order construct. For example, ceteris paribus, greater will be positively related to the treatise’s eventual novelty enhances the opportunities for subsequent perceived importance among management scholars. scholarship on new and unexplored questions and, thus, extendibility. Simultaneously, as noted above, Scientific Rigor extendibility breeds further novelty. Similarly, Besides the meaningfulness of its substance and practical relevance enhances extendibility, and vice- insights, the scientific rigor/validity of a theory/ versa: greater practical relevance facilitates direct model is a conventional assessment criterion observation of the relevant phenomena as well as emphasized by prior theory-building authors of manifestations of the theoretical knowledge (Arend et al., 2015; Bacharach, 1989; Dubin, “in use,” which may provide new ideas/insights to 1969; Eisenhardt, 1989b; Shapira, 2011) as well as modify/extend the theory. At the same time, greater by editors of leading publications (e.g., Bartunek extendibility magnifies the practical relevance of the et al., 2006; Whetten, 1989), and which has been theory as it broadens its applicability to a greater related to the prominence of theoretical works in number of phenomena, contexts, or problems of management (Colquitt and Zapata-Phelan, 2007; consequence to practitioners. Miner, 2003). Drawing from philosophy of science In short, we expect complex interdependence precepts, two rigor/validity aspects in particular among the dimensions associated with the have been emphasized in the prior literature as explanatory significance and usefulness of a necessary requirements for sound theoretical theoretical contribution, so that their discrete construction: (1) logical consistency (Ackoff, 1962; impact on the attractiveness of the contribution to Campbell, 1953; Oxley et al., 2010; Shapira, 2011) scholars will be difficult to untangle. Consequently, and (2) empirical falsifiability (Bacharach, 1989; we argue that their impact will be best captured Popper, 1959; Shapira, 2011). through a comprehensive, unified dimension. We define logical consistency as the extent to which We thus propose the notion of explanatory the theoretical contribution is a well-constructed meaningfulness of a theory as the overarching deductive system resulting in coherent explanation explanatory value perceived by management and concluded insights. Namely, starting with clear scholars based on a theory’s combination of novelty, construct definitions as well as explicit statement of extendibility, and relevance to practice. Moreover, assumptions or axioms and, based on these, solid we propose that this overarching construct can be logical derivations follow which lead to specific best operationalized as a second-order factor with predictions (Ackoff, 1962; Campbell, 1953; Dubin, novelty, extendibility, and relevance as its three 1969; Haveman et al., 2019; Shapira, 2011). reflective lower-order indicators. Formally: Logical consistency, therefore, implies that there are Hypothesis 1: Perceived novelty, extendibility, and no tautologies or near-tautologies in the arguments relevance to practice will be distinct yet interrelated (Arend et al., 2015; Bacharach, 1989; Priem & constructs whose covariance will be effectively captured Volume 20, Number 2 • April - June 2020 81

Butler, 2001) and that the logic is otherwise reliable It is important to note that, in contrast to and coherent (Ackoff, 1962; Campbell, 1953; the hard sciences, falsifiability of organization Dubin, 1969; Shapira, 2011; Whetten, 1989). and management theories is obstructed by (i) Since logical consistency has been argued to be an the complex, open, and changing nature of important aspect of “good” theory, we expect this organizational phenomena; (ii) the inherent trait to increase the attractiveness of a theory to vagueness of our theories, which rarely can be management scholars and, thus, their allegiance to specified in mathematical terms; and (iii) the it, leading to eventual impact. Hence: pervasive lack of clarity regarding assumptions and Hypothesis 3: The logical consistency of a theoretical boundary conditions (Acar et al., 2020; Bacharach, treatise will be positively related to its eventual 1989; Miller & Tsang, 2010). Therefore, strict perceived importance among management scholars. falsifiability leading to definitive confirmation/ We define the empirical falsifiability of a theoretical refutation is an ill-fitting notion and thus an contribution as the extent to which predictions unreasonable demand on management theories, as formulated from it can be refuted by empirical well as on theories in the social sciences in general evidence (Arend et al., 2015; Bacharach, 1989; (Acar et al., 2020). Rather, falsificationism in the Miller & Tsang, 2010; Shapira, 2011).6 Empirical social sciences needs to be understood as a more falsifiability is related to the ability to operationalize modest pursuit, where verification and falsification and measure constructs accurately and reliably; to claims should be regarded as tentative or possibly the precision of predictions regarding the nature “fallible” (Miller & Tsang, 2010). Still, seeking of the relationships among them as well as the to plausibly affirm or reject theories on the basis specifications of boundary conditions under which of rigorous empirical tests is essential for the such relationships should hold; and to the ability to production of knowledge in organization science obtain pertinent data, including enough variation (Miller & Tsang, 2010; Oxley et al., 2010). in the object of analysis (Astley & Zammuto, Given the emphasis in the theory-making/ 1E9X9P2L; ABNaAcThIaOrNach&, R1IG9O89R)I.NFTaHlsEiOfiaRbIiZlIiNtyG therefore theory-evaluation literature upon practicable facilitates empirical tests of the theory, as well as testability leading to a sense of empirical validity o3f9 the derivation of conclusions from such tests, thus management theories (Bacharach, 1989; Eisenhardt, contributing to cumulative learning. 1989b; Miller & Tsang, 2010; Oxley et al., 2010; Shapira 201), we expect empirical falsifiability to be Figure 1: Hypothesized Model of Theory Importance FIGURE 1: Hypothesized Model of Theory Importance 82 Journal of Management Research

one of the determinants of a theory’s attractiveness exemplar, respondents could opt to either end to scholars, contributing to its use (McKinley et al., the survey or select a second article form the list 1999) and, thus, to its eventual prominence in the and complete a second assessment. To ensure that field (Miner, 2003). Therefore: respondents only rated the theoretical exemplars Hypothesis 4: The empirical falsifiability of a they were most familiar with, the survey closed after theoretical treatise will be positively related to its the second assessment round. eventual perceived importance among management Of 2,918 scholars invited to participate, 641 scholars. entered and completed the survey, for a 22% Our hypothesized model is depicted in Figure 1. We response rate. Among those that saw the survey now turn to the description of our empirical test of to its end, 187 respondents reported insufficient this model. familiarity to provide a detailed assessment of any of the listed theoretical exemplars, 153 respondents METHODS completed the assessment instrument for only one exemplar, and 301 assessed two exemplars, Participants, Procedures, and Study Sample producing a total of 755 theory assessment Participants in our study were academic members observations. Of these, 137 assessments included of the Business Policy and Strategy (BPS), an “I don’t know” response to one (or more) Organization and Management Theory (OMT), evaluation item(s) and were dropped from the and Entrepreneurship (ENT) divisions of the present study9 in order to limit the sample to Academy of Management who were affiliated with entirely knowledgeable raters. To further screen for institutions within the United States.7 Potential expert assessments, we also dropped an additional respondents were contacted by email and invited 106 observations where the respondent had rated to participate in an online survey that entailed his/her familiarity with the subject article below 5 a detailed assessment of up to two theoretical in a 7-point Likert scale. These screens resulted in a exemplars, chosen from a short list of eight well- study sample of 512 observations. known works in strategy and organization theory. Among the participants comprising the final To promote response, the articles offered for sample, 38% were members of the BPS division assessment were seminal pieces within four of only, 19% were OMT members only, 9% were the most prominent theoretical streams within ENT members only, 32% were members of two of strategic management and organization theory. At these divisions, and 1% reported membership in all the same time, to provide variance in the variables three. Over 97% held doctorates and had had this of interest, they represented different theoretical degree for an average of 13 years. After screening approaches and/or different levels of influence for this, the average reported familiarity with the within their respective literatures. The works offered theories evaluated was 6.04 out of 7, and 94% of as assessment choices were Jensen and Meckling sample respondents reported using the assessed (1976) and Eisenhardt (1989a) for agency theory, article at least once as a reference in their own work. DiMaggio and Powell (1983) and Greenwood and Hinings (1996) for institutional theory, Barney Measurement (1991) and Amit and Schoemaker (1993) for the Scholarly importance. Following Miner (2003), resource-based view, and Williamson (1975) and our dependent variable was measured using a Ouchi (1980) for transaction-cost economics. single-item indicator of the management scholars’ At the start of the survey, participants were asked perceived importance of the theoretical exemplar. to select the theoretical work that they were most Respondents were asked to “rate the importance of knowledgeable about to assess.8 After completing the theory to the field of management,” and their the assessment instrument for the chosen theoretical responses were recorded on a 7-point scale ranging Volume 20, Number 2 • April - June 2020 83

from “not at all important” to “very important.” (EFA). Scores on the 30-item battery from the The validity of this measure was explored by study sample were factor analyzed using principal contrasting it with citation data from the EBSCO components extraction and oblique (promax) Business Source Complete database: The correlation rotation. Five factors were initially retained based of average perceived importance with total number on the Kaiser-Guttman rule (Kaiser, 1960) and of citations was .75, and that with average annual the scree test (Cattell, 1966). Following rotation, citations was .73, which supports the convergent five items were eliminated which either had validity of our measure. cross-loadings or failed to load onto any factor. A Theory evaluation constructs. Since novelty, subsequent factor analysis on the remaining 25-item extendibility, relevance to practice, logical instrument, using the same methods and retention consistency, and empirical falsifiability had not been criteria, resulted in five factors that accounted operationalized via survey instruments before, we for 64.6% of item variance. The extracted five- developed a theory evaluation scale to measure factor structure was consistent with our proposed the independent variables in our proposed model. constructs, and there were no items with cross- Design of this instrument was based on our review loadings above .35 on the rotated factors. Loadings, of the theory building/evaluation literature in item communalities, percent variance explained by management and was guided by the conceptual each factor, and Cronbach’s a reliability estimates of definitions offered earlier in this paper. It also the final subscales are presented in Table 1. underwent two separate phases of refinement before The resulting measures from the validated 25-item being used to test the hypotheses in this study. theory evaluation instrument are as follows: The In the first phase, different versions of our proposed novelty of the insights proposed by a theoretical theory evaluation scale underwent sequential review exemplar was measured using a six-item scale by two separate panels of judges: A first draft of (a = .89). Sample items are: “the theory offered the subscales for each theory evaluation construct different explanations for previously established was subjected to scrutiny by three professors from relationships” and “the theory created an ‘aha’ sociology and management departments who moment when proposed.” The extendibility of the were knowledgeable about the relevant literature, constructs and rationale presented in the theoretical and who reviewed for content validity and clarity. exemplar was operationalized via a four-item scale Items were dropped, added, or amended based on (a = .73). Examples of items are: “the theory as the feedback received, resulting in a preliminary presented could be extended to a broad range of 31-item overall theory evaluation scale. The latter, phenomena” and “the theory’s rationale could be along with other items capturing variables to be extended to shed light on new questions.” Relevance used as screens or controls in the study, was then to practice was measured using an eight-item scale reviewed by a larger panel of faculty and doctoral (a = .91). Items include: “understanding the theory students, who focused on (i) the clarity and interest helps practitioners formulate action plans” and “the of the questions, (ii) appropriateness in terms of theory offers a good tool for business consultants.” the time and intellectual demands on respondents, The logical consistency of the theoretical exposition and (iii) sequencing of the different parts of the was captured with a three-item scale (a = .63), survey instrument. Besides adjustments to other including “there were inherent contradictions in survey items, the theory evaluation scale underwent the theory” and “the propositions derived from further editing, and one of the preliminary items the theory were tautological,” both reverse-scored. was altogether dropped. Finally, the empirical falsifiability of predictions The second phase of scrutiny consisted of validation derived from the theory was measured using a four- and refinement of the factor structure of the theory item scale (a = .85) that includes items such as evaluation scale through exploratory factor analysis “predictions from the theory can be contrasted with empirical evidence” and “the theory’s constructs 84 Journal of Management Research

Table 1: Exploratory Factor Analysis of (Refined) Theory Evaluation Scalea Factor pattern (standardized regression coefficients) Item Mean s.d. Relevance to practice Novelty Falsifiability Extendibility Logical consistency h 2 P1 5.26 1.68 0.88 0.01 0.04 0.03 -0.12 0.79 P2 5.34 1.65 0.84 0.08 0.02 -0.02 -0.06 0.74 P3 5.26 1.58 0.74 -0.12 0.02 -0.13 0.29 0.61 P4 5.52 1.47 0.76 0.11 0.09 0.00 0.02 0.73 P5 5.22 1.66 0.90 -0.10 -0.05 0.01 -0.07 0.71 P6 6.31 1.01 0.42 0.14 -0.04 0.31 0.12 0.53 P7 5.37 1.58 0.67 -0.04 -0.08 0.02 0.21 0.51 P8 5.46 1.59 0.66 0.13 0.15 0.03 -0.07 0.63 N2 5.74 1.42 -0.05 0.82 0.10 -0.16 -0.03 0.61 N3 5.92 1.29 -0.01 0.82 0.10 -0.07 -0.05 0.66 N4 5.97 1.31 -0.03 0.84 0.03 -0.09 -0.06 0.61 N5 6.18 1.21 0.02 0.79 -0.04 0.10 0.02 0.70 N6 6.28 1.07 0.10 0.82 -0.09 0.06 -0.01 0.73 N7 5.71 1.46 0.01 0.70 -0.04 0.10 0.12 0.62 F2 5.41 1.57 0.00 0.08 0.71 0.08 0.21 0.77 F3 5.52 1.49 -0.01 0.07 0.69 0.09 0.23 0.76 F4 4.62 1.79 0.03 -0.04 0.89 0.03 -0.04 0.77 F5 5.46 1.54 0.09 0.01 0.74 -0.02 -0.10 0.55 E1 5.93 1.31 0.03 0.29 0.03 0.61 -0.04 0.62 E2 5.02 1.83 0.11 -0.17 0.13 0.70 -0.19 0.50 E3 5.81 1.43 -0.10 -0.15 0.09 0.86 0.01 0.65 E5 6.21 1.12 -0.02 0.13 -0.07 0.68 0.04 0.53 LC1 5.54 1.72 0.02 0.11 -0.18 0.34 0.54 0.56 LC2 4.65 1.75 0.17 -0.10 0.01 -0.13 0.76 0.57 LC3 4.72 1.89 -0.12 0.06 0.20 -0.06 0.76 0.69 Total communality 16.15 % Variance explained 19.10 17.01 10.76 10.21 7.50 64.58 Alpha coefficient 0.91 0.89 0.85 0.73 0.63 a n = 512. Promax-rotated factor solution reported. Factor loadings (i.e., coefficients ≥.35) are indicated in boldface. Interfactor correlations are in the range from 0.26 to 0.43. h2 = item communalities. Percent variance explained statistics are reported from the pre- rotation varimax solution, since, once factors are allowed to covary with the additional procrustean transformation, variance explained overlaps and communality cannot be uniquely partitioned. are easily translated into measurable variables.” All Control variables. To minimize potential omitted measures used a Likert response scale, ranging from variable problems when testing for hypothesized 1 = “strongly disagree” to 7 = “strongly agree.” The relationships to perceived theory importance, we complete text of the theory evaluation instrument is controlled for theory age, as well as for possible reproduced in the Appendix. rater biases due to years of experience, closeness to the theory, and respondent research and consulting Volume 20, Number 2 • April - June 2020 85

intensity. Theory age was measured as the logarithm and five first-order factors, where only three of them of the number of years since publication of the (novelty, extendibility, and relevance to practice) evaluated article. Rater experience was measured as loaded on the second-order factor. Finally, Model log (1 + number of years since respondent earned 5 was an alternative structure with one general his/her doctoral degree). Rater familiarity was second-order factor with all five first-order factors measured as the logarithm of the reported number loading onto it.10 The comparison among the last of instances of use of the rated article as a reference three models served as the test of Hypothesis 1. in the rater’s own publications or presentations. To gauge and compare model fit, five recommended Research intensity and consulting intensity were the measures were used: the chi-square/df ratio; the root self-reported percentages of the rater’s professional mean square error of approximation (RMSEA); life spent in research and consulting activities, the non-normed fit index (NNFI), also called the respectively. Tucker-Lewis index (TLI); the comparative fit index (CFI); and the consistent Akaike information Analyses criterion (CAIC) (e.g., Bentler, 1990; Kline, 1998). We used covariance-based structural equation This is a comprehensive set that includes absolute, modeling (SEM) with LISREL software. Following incremental, and parsimony fit indices. Chi- recommended procedures (e.g., Anderson & square/df is a measure of absolute fit, indicating Gerbin, 1988), we employed a two-step analytical how closely the estimated covariances in a fitted strategy whereby, prior to fitting the structural model match the covariances in the original data. model, we examined and determined the best Normally, values of 3.0 or less in this statistic measurement model through confirmatory factor are interpreted as indicating good fit to the data analysis (CFA). In our case, this step also served to (Kline, 1998), although others have proposed a test our first hypothesis. All latent variables were cut-off of 5.0 or less as indicative of acceptable fit modeled using the discrete item indicators from (Wheaton et al., 1977). RMSEA is another absolute our validated instrument (as opposed to using item fit index which is adjusted for sample size and thus averages), and models were fit using the maximum recommended to complement the chi-square fit test likelihood estimator, which has been found to be in larger samples. A value smaller than .08 in this rather robust to distributional variations (e.g., Chou index is considered to indicate good fit (MacCallum & Bentler, 1995). et al., 1996). RMSEA also has the advantage of a We compared five models to identify the known distribution, allowing us to also report the measurement structure that best fit the data. Model 90% confidence interval around point estimates. 1 was a first-order unitary factor model, with all By contrast, NNFI and CFI are incremental fit items loading onto a single latent variable. Model indices, indicating improved overall fit relative to 2 was a first-order model with the five orthogonal a null model where all variables are assumed to factors, and with the loading structure as derived be uncorrelated. For both indices, values equal from the EFA. The comparison between these to or exceeding .95 are indicative of good fit (Hu two models would serve to confirm the multi- & Bentler, 1999). Finally, CAIC is a parsimony dimensionality of our theory evaluation instrument. fit index (i.e., it penalizes for model complexity) Model 3 was the correlated (oblique) five-factor that also adjusts for sample size. Because it is not model. The comparison between Model 2 and bounded, there is no suggested cut-off for this Model 3 therefore would serve to confirm that there statistic. Rather, the model with the lower value in is significant covariance between the first-order this index is seen as superior (Akaike, 1974). factors which, as hypothesized, might then best be Beyond fit index statistics, we also used chi-square captured through a second-order factor. Model 4 difference tests between nested models to assess was the hypothesized measurement model with one which of these provided a marginally better fit. second-order factor (explanatory meaningfulness) 86 Journal of Management Research

Models 2 and 3 are nested models, and Model 5 is in Model 3 were significant at p < .001, and the also nested within Model 4. average estimated factor correlation was quite high Once the best measurement model had been (0.53), although well below 0.9 in all cases11 thus identified, SEM was performed to estimate the providing evidence of the discriminant validity fit of the overall hypothesized model to the data. of our constructs (e.g., Venkatraman, 1990). Beyond overall fit, path coefficients for proposed These results lend strong support to our theory relationships were used to test the remaining evaluation constructs being distinct, yet covariant, hypotheses. dimensions. The high correlations among the first- order constructs boded well for second-order factor RESULTS models. Model 4, our hypothesized model with a three- Measurement Model dimensional second-order meaningfulness factor, Table 2 shows the fit indices for the alternative demonstrated marginally better fit than Model 3 measurement models. Model 1 exhibited a rather (c2/df = 3.65; RMSEA = .075; NNFI = .96; CFI poor fit to the data (c2/df = 11.41; RMSEA = .18; = .96; CAIC improving to 1,445). Thus, Model 4 NNFI = .83; CFI = .84). Model 2, with the five- is preferred, largely on the grounds of parsimony factor structure as revealed by the EFA, produced (i.e., it provides as good a fit while involving fewer substantially better fit indices, although not yet parameter estimates). Moreover, all parameter demonstrative of good fit (c2/df = 6.02; RMSEA = estimates in Model 4 were strongly significant (p < .11; NNFI = .92; CFI = .92; CAIC = 2,417 versus .001 in all cases), including the loadings of novelty, 5,361 for Model 1). Therefore, the comparison of extendibility, and practical relevance onto the fit indices between these two models confirms the second-order construct, which were .77, .77, and multidimensionality of our theory evaluation scale. .64 respectively. Model 3, with the five correlated factors, Finally, Model 5, where the variances among all five demonstrated good fit to the data (c2/df = 3.66; first-order factors were explained by one general RMSEA = .075; NNFI = .96; CFI = .96; CAIC second-order factor, produced a slightly worse dropped to 1,460). Moreover, the chi-square fit than either Model 3 or Model 4 (c2/df = 3.68; difference test between Models 2 and 3, c2diff.(10) RMSEA = .075; NNFI = .96; CFI = .96; CAIC = = 686, p < .001, strongly suggests that the latter 1,459). Also, a nested comparison between Model is preferred. Additionally, all parameter estimates 4 and Model 5 shows that the former (with only novelty, extendibility, and practical relevance Table 2: Fit Indices of Alternative Measurement Modelsa Model χ2 df χ 2 /df RMSEA (90% CI) NNFI CFI CAIC Model 1 3,138.8 275 11.41 .18 .83 .84 5,361 Model 2 1,654.7 275 6.02 .11 .92 .92 2,417 Model 3 265 3.66 .075 (.070 ̶ .080) .96 .96 1,460 Model 4 968.7 269 3.65 .075 (.070 ̶ .079) .96 .96 1,445 Model 5 980.7 270 3.68 .075 (.071 ̶ .080) .96 .96 1,459 994.3 a n = 512. Model 1 has one first-order unitary factor. Model 2 has five orthogonal first-order factors. Model 3 has five correlated (oblique) first-order factors. Model 4 has five first-order factors and one second-order factor (Explanatory Meaningfulness) with only three of the first-order factors (Novelty, Extendibility, Relevance to Practice) as its underlying dimensions. Model 5 has one general second-order factor with all five first-order factors as its underlying dimensions. 90% CI = 90% confidence interval around RMSEA point estimate. Volume 20, Number 2 • April - June 2020 87

loading on the second-order factor) is strongly and as is the default in LISREL. The hypothesized preferred: c2diff.(1) = 13.6, p < 0.001. In other model exhibited good fit to the data (c2/df = 2.11; words, the hypothesized measurement structure RMSEA = .06; NNFI = .95; CFI = .96), and provided better fit than alternative second-order it explained 86% of the variability in perceived models. importance. The standardized second-order loadings Overall, these results support the notion that and path coefficient estimates in the fitted model novelty, extendibility, and practical relevance are are shown in Figure 2. distinct dimensions, but also collectively reflective Hypothesis 2, which postulates that the explanatory of a second-order construct, which we have termed meaningfulness of theoretical exemplars will the explanatory meaningfulness of a theory or model. be positively related to the eventual perceived Thus, Hypothesis 1 is supported. importance by management scholars, was supported by a positive and strongly significant Structural Model path coefficient between these variables (b = After finding support for our measurement model, 1.42, p < .001). By contrast, Hypothesis 3, which we proceeded to test the complete hypothesized predicts that the logical consistency of theoretical structural model of theory importance, as depicted exemplars will also be positively related to eventual in Figure 1. To provide a rigorous test of our scholarly importance, was not supported, as the hypotheses, we also included the control variables path coefficient was actually negative although not d(eEixsXcpPulLasAnseNadtAoTryIaObNmov&eea.nRIiGnTgOhfRuelInNeisTnsH,dEelOpoeRgniIcZdaIeNlnGtconvsaisrtieanbcleys, statistically significant (b = – .57, p = .12). Our and empirical falsifiability) were allowed to covary, eresstiumltasteadlsocoefaffiilciteontsufrpopmoretmHpiyrpicoatlhfeaslissifi4a,bialis4ty2thtoe as indicated by our test of the measurement model, scholarly importance was not statistically significant either (b = – .15, p = .35). a an FIGUFiRguEre2:2:SStrtruuccttuurraallEEqquautaiotinonMoMdoeldineglinRgesRuletssaultsa *n=** =5*pp152≤≤1p.2..S10≤.t1aS.n1tdaanrddaizreddizeestdimeastteismreapteorsterde.pHorytpeodt.hHesyizpeodtrheelastiizoendshripesladtiisopnlasyheidpsindbisopldla. yed in bold. *** *p* ≤p.0≤0.10,1two-tailed. *** p ≤ .001, two-tailed. 88 Journal of Management Research

DISCUSSION with the proposed theory evaluation constructs of This paper synthesizes the prior literature on theory novelty, extendibility, practical relevance, logical building and evaluation in management to propose consistency, and empirical falsifiability. a model of the key aspects of a theoretical treatise In turn, CFA findings support the existence of that drive its eventual scholarly importance. In the hypothesized second-order theory evaluation our model, the novelty, a priori extendibility, and factor, with novelty, extendibility, and practical practical relevance of a theory’s elements capture relevance as its underlying dimensions, and thus its explanatory value and lead to its adoption, are consistent with our proposed explanatory allegiance, and eventual salience among scholars. meaningfulness construct. Finally, SEM findings Moreover, we propose that these aspects are show this explanatory meaningfulness factor to interconnected and mutually reinforcing, so that be a statistically significant, and very important, they are best understood as subdimensions of an predictor of the perceived current importance of overarching construct, which we have termed a theoretical article to the field of management. the explanatory meaningfulness of the theoretical By contrast, after controlling for explanatory work. Therefore, we hypothesize that novelty, meaningfulness, the logical consistency and extendibility, and practical relevance are the empirical falsifiability properties of theoretical reflective indicators of a second-order latent works are not significantly related to their perceived variable, and that this higher-order variable will current importance. be positively related to the subsequent perceived importance of theoretical exemplars among Research Implications management scholars. Our findings have implications for the theory Alongside explanatory value, the scientific rigor of making/evaluation literature in management. the exposition is also seen in most of the theory In particular, they inform the ongoing debate making/evaluation literature in management between those that emphasize the explanatory role as either an essential requirement or, at least, a of theories (Astley & Zammuto, 1992; Corley & desirable trait of productive theory. Therefore, we Gioia, 2011; McKinley et al., 1999; Sutton & expect that rigor will also impact the prominence Staw, 1995) and the importance of novel, rich, and of theoretical works. In particular, we hypothesize relevant accounts of multifaceted organizational that the logical consistency and empirical falsifiability realities (Davis, 1971; Eisenhardt, 1989b; Poole of theoretical exemplars will be positively related & Van de Ven, 1989) and those that insist on the to their eventual perceived importance among importance of rigorous construction and empirical management scholars. refutability of theoretical frameworks (Arend et al., Our model was tested with data from members 2015; Miller & Tsang, 2010; Oxley et al., 2010; of the OMT, BPS, and ENT divisions of the Shapira, 2011). Academy of Management, who provided a detailed An important undercurrent in this debate stems assessment of one or two seminal theoretical from the existence of trade-offs between explanatory exemplars with which they purported to be very and scientific rigor aims. For example, there is a familiar. Respondents were asked to rate evaluatory trade-off between extendibility through broadly traits of the particular theoretical treatise in its defined boundary conditions and constructs, on the original published form, as well as to assess its one hand, and the precision of predictions and thus current importance as a contribution to the field of falsifiability of a theory, on the other hand (Astley management. EFA findings support the convergent & Zammuto, 1992; Bacharach, 1989; Makadok et and discriminant validity of our measurement al., 2018). Also, construct vagueness in particular scales across a structure composed of five distinctive is deplored by those who emphasize theoretical but correlated dimensions, which are consistent rigor and point to its unfortunate consequences, Volume 20, Number 2 • April - June 2020 89

including muddiness of thought and infinite Does this mean that scientific rigor plays no regress problems (e.g., Ackoff, 1962; Hallberg & relevant role in the minds of most management Felin, 2020; Kaplan, 1964; Turner, 1989), as well scholars and can be disregarded by those seeking as obfuscated empirical testing (Miller & Tsang, to make lasting and influential theoretical 2010; Oxley et al., 2010). Similarly, there is a contributions? We do not believe the findings trade-off between the construction of encompassing from the present study warrant such a radical theories through the use of paradox and the pursuit conclusion. Unless a theoretical exposition is of logical consistency (Poole & Van de Ven, 1989; minimally bounded by (even implicit) assumptions, Smith & Lewis, 2011). is based on reasonably well-defined constructs There has been a paucity of empirical work and a coherent rationale, and provides a frame contrasting the above perspectives and shedding from which to deduct fairly explicit and testable light on some of the extant debates; hence the predictions, it is unlikely to meet the requirement relevance of the present work. Results from this for publication in management journals in the study suggest that the perceived explanatory first place (e.g., Bartunek et al., 2006). Likewise, significance and usefulness of a theory drives a measure of theoretical rigor will be necessary for its eventual prominence within the research subsequent tests of the theory to lead to reasonably community in organization theory and strategic unequivocal knowledge claims, thus generating a management, irrespective of its degree of scientific sense of constructive knowledge accumulation from rigor. Management scholars appear to prize it. This, in turn, will underpin persistent concord contributions that provide meaningful explanation on the theory’s usefulness and, thus, its longevity. (through greater novelty, extendibility, and practical In short, a minimum level of rigor is an obvious relevance), while not much weight is being put on necessary condition for a management theory to be the greater logical consistency and falsifiability of seen as a valid and beneficial vehicle for knowledge such treatises. production, and hence for it to be adopted and As such, our findings support Davis’s (1971, p. used. 309) classic assertion for the social sciences that Rather than rigor playing no role in the subsequent “a theorist is considered great, not because his adoption and prominence of management theories, theories are true, but because they are interesting. we interpret the results from the present study to . . . In fact, the truth of a theory has very little to suggest that, beyond acceptable levels, greater do with its impact,” or McKinley et al.’s (1999, p. logical consistency and empirical falsifiability 636) proposition that “empirical validity tends to do not contribute to greater appreciation of the recede into the background as a determinant of theoretical work by management scholars. In other where organization theorists place their scholarly words, we speculate that scientific rigor matters for allegiances.” Our findings also support those scholarly impact only up to a threshold level. Once who advocate for theory-making that embraces this threshold of acceptability is met, additional inconsistencies, tensions, or contradictions rather formality and/or precision of argument do not than simplifying them away (e.g., Lado et al., breed additional impact within the management 2006; Poole & Van de Ven, 1989; Smith & Lewis, academic community, at present. 2011). What is sought are models that offer new The threshold level of rigor where effects on impact language, perspective, rationale, constructs, and/ are exhausted may also evolve with the state of or relationships, which can be used to improve development of a particular research literature. In scholars’ ability to make sense of complex business the initial stages of scientific inquiry into a new realities and derive insights on a broad range of domain, vague conceptual frameworks may be important organizational phenomena, as well needed before models and, eventually, theories as help practitioners identify, characterize, and can be built (Oxley et al., 2010; Priem & Butler, interpret such phenomena in the real world. 90 Journal of Management Research

2001; Shapira, 2011). To the extent that this is the For those who advocate for greater rigor in case, our focus on seminal works in the present management theory-making (e.g., Acar et al., 2020; study may have underplayed rigor effects: i.e., our Oxley et al., 2010), our results provide a measure exemplars may have required low thresholds of of the challenge ahead: The current incentives and rigor at the dawn of their conceptual streams to be norms in the field appear to run counter to their considered important theoretical contributions. In aspirations for more consistently reliable knowledge other words, despite significant variation in rigor claims, as well as more productive accumulation of in our data, the positive effect of rigor on impact a coherent body of understanding of organizational (below the acceptability threshold) may have been phenomena. At present, logical consistency and censored due to a lack of variation in the stage of precision of prediction appear to receive (at best) theoretical life cycle among the exemplars used. limited appreciation from management scholars. Interestingly, our results, as well as the above Our research also has direct implications for notion of truncated rigor effects on impact, lead theorists, as well as for editors and reviewers to an alternative interpretation of past findings in engaged in theory evaluation in management. The the incipient empirical literature on the subject. recommendation for theorists is straightforward: In particular, Colquitt and Zapata-Phelan (2007) our results underscore the importance of found that (i) articles that introduced new advancing constructs, rationales, and hypothesized mediators or moderators of existing relationships, relationships that are novel, extendible, and and which relied on loose conceptual arguments practically relevant, as this will increase the or (at best) graphical models or diagrams (i.e., meaningfulness of the resulting theory to medium level of rigor), as well as (ii) articles management scholars, which in turn will drive that either introduced new relationships among it to prominence over time. Still, theory writers existing constructs or reconceptualized constructs, should avoid novelty that emanates from a lack and which relied on either models/diagrams or of awareness of prior work and thus results in “true theory” (i.e., higher level of rigor), were the “reinventing the wheel” without really making most impactful. From this, they interpreted that a substantial contribution. Writers should also the articles that achieve the greatest impact in avoid extendibility that emanates from unjustified management are those that offer “a balance between vagueness of constructs, rationale, and boundary novelty and continuity” with existing theoretical conditions, including from unnecessarily obscure formulations (p. 1293). The present study suggests and convoluted writing style (e.g., Tourish, 2020). a plausible more parsimonious explanation for Given the trade-offs between precision and meaning, Colquitt and Zapata-Phelan’s findings: Simply, as well as the limited impact effects of rigor, that the most cited articles are those that utilize management theorists are also encouraged not to shy the minimum necessary theoretical rigor while away from the embryonic exploration of important maximizing the richness and flexibility of their new phenomena and rationales, or to compromise newly introduced elements. As noted above, there the potential broadness of their ideational scope are trade-offs between the precision of prediction and/or the multifaceted richness of their account, in that comes with greater theoretical rigor and exchange for excessive exactitude of their derivation. the explanatory meaningfulness of the account. After meeting reasonable requirements for clarity, Therefore, the most impactful articles may be those coherence, and precision, in correspondence with the that meet the thresholds of logical consistency and stage of conceptual development of the domain as falsifiability required by the stage of development of well as the current norms in the field, increasing the their literature, but no more, in exchange for greater formal elegance and meticulousness of the exposition broadness and richness of their exposition. is unlikely to increase the appreciation of their work. Volume 20, Number 2 • April - June 2020 91

This implies that the skill set for impactful theory First, we used seminal works as vehicles for data writing in management today is very different from, collection, which, as discussed above, might have say, in economics or finance. Impactful theory contributed to the censoring of scientific rigor building may also be particularly challenging for effects. Subsequent investigations could further budding management theorists, as they must be validate our model by relying on rater assessments willing to leave behind the “safety” of an airtight of different sets of less foundational conceptual formulation to, for example, accept the vagueness works. Second, our study focused on a few research associated with theorizing about incipient but streams in organization theory and strategy, relevant organizational phenomena and managerial which might have introduced biases in our data. concerns. Although (in further analyses not reported here) the As to editors and reviewers, we hope that our relationship between explanatory meaningfulness synthetic model of theory evaluation dimensions and importance was found to be stable across each (novelty, extendibility, relevance, internal of our four theoretical stream subsamples, it would consistency, and falsifiability) may help consolidate be important to investigate if findings generalize and organize the core set of assessment criteria across other theoretical streams and management for theoretical treatises, alongside other essential disciplines. In particular, our focus on macro- considerations like familiarity and interconnection level theories might be complemented with similar with the prior literature or writing quality. With studies using micro-theories and respondents from regard to the weight given to different criteria, associated disciplines (i.e., in human resource findings from our study suggest that richer and management and organizational behavior). more encompassing theories that necessarily Likewise, future research could explore how drivers give away some rigor and coherence of argument of importance may differ across disciplines with are likely to achieve greater salience. Thus, our disparate theorizing traditions (e.g., organization findings may be consistent with recent editorial science versus operations research and management calls to relax conventional deductive theory- science). Third, as this is the first attempt at testing making requirements so as to broaden the gamut of our parsimonious model, our study was confined accepted approaches to theorizing (e.g., Haveman to American researchers in order to facilitate et al., 2019). Still, we believe that the apparent lack follow-up during data collection, as well as to of impact effects from greater rigor in the present control for likely confounding issues arising from study should give pause to those associated with differing cultural contexts and research traditions. management journals. Journals play an important Nevertheless, it would be important to extend tests role in setting the incentives and quality norms of the model to other cultural settings. It is thus for research in the discipline. Arguably, a greater hoped that this study will generate a stream of appreciation for clean and tidy logical deduction systematic tailored investigations, as well as cross- and for precise prediction in management theories national contrasts. would contribute to more productive knowledge The discovery in this study of the higher-order creation and accumulation going forward. Thus, construct of explanatory meaningfulness also opens further scientific maturity of the management up promising avenues for further research. We field may require a proactive recalibration of the provide preliminary evidence of its nomological importance afforded to rigor in theory evaluation validity in the form of its relationship to policies and norms. perceived scholarly importance, but other aspects of its extended nomological network could be Limitations and Future Directions further explored. For example, is explanatory As in any empirical study, our research design meaningfulness also related to the application and involved trade-offs whereby advantages were influence of a theory among practitioners? Likewise, obtained at the expense of some limitations. the relationship between various aspects of scientific 92 Journal of Management Research

rigor in theorizing and the scholarly prominence evaluation dimension reflected in a theory’s novelty, of theoretical treatises in management deserves extendibility, and practical relevance) and provide further inquiry. Is there no relationship, a truncated empirical evidence of its manifestation, as well as relationship (as speculated here), or even a quadratic of its being a strong and positive predictor of long- relationship whereby higher levels of rigor actually term scholarly importance. detract from subsequent impact? Our findings point to clear recommendations for theory building and evaluation in management. CONCLUSION Specifically, given current norms in the field, our This article makes several contributions to the study suggests that theory writers might want to sparse empirical literature on theory evaluation err on the side of lesser formality and precision in dimensions, and on how the latter relate to the their formulation in exchange for greater originality, eventual prominence of theoretical works in relevance, and multifaceted richness (i.e., meaning) the organization and management field. Most of their account. The model and results presented importantly, we propose and test a comprehensive here pave the way for subsequent conceptual but parsimonious model of attributes that drive and empirical research on drivers of scholarly the scholarly importance of a theoretical treatise. impact of theoretical contributions in the field of We also introduce the explanatory meaningfulness management. A fuller understanding of the etiology construct (a tripartite second-order theory of management theories’ impact should be an important and, hence, welcome pursuit. REFERENCES Acar, W., Franquesa, J., & Mwaka, J. O. (2020). Conceiving perfectible theories in management through adaptive framing. International Journal of Strategic Decision Sciences, 11(1), 1–20. Ackoff, R. L. (1962). Scientific method. New York: Wiley. Akaike, H. (1974). A new look at the statistical model identification. IEE Transactions on Automatic Control, 19(6), 716–23. Amit, R., & Schoemaker, P. J. H. (1993). Strategic assets and organizational rent. Strategic Management Journal, 14, 33–46. Anderson, J. C., & Gerbing, D. W. (1988). Structural equation modeling in practice: A review and recommended two step approach. Psychological Bulletin, 103, 411–423. Arend, R. J., Sarooghi, H., & Burkemper, A. (2015). Effectuation as ineffectual? Applying the 3E theory-assessment framework to a proposed new theory of entrepreneurship. Academy of Management Review, 40(4), 630–651. Astley, W. G., & Zammuto, R. F. (1992). Organization science, managers, and language games. Organization Science, 3, 443–460. Bacharach, S. (1989). Organizational theories: Some criteria for evaluation. Academy of Management Review, 14(4), 496–515. Barney, J. B. (1991). Firm resources and sustained competitive advantage. Journal of Management, 17(1), 99–120. Barney, J. B. (2018). Editor’s comments: Theory contributions and the AMR review process. Academy of Management Review, 43(1), 1–4. Bartunek, J. M., Rynes, S. L., & Ireland, D. (2006). What makes management research interesting, and why does it matter? Academy of Management Journal, 49(1), 9–15. Bentler, P. M. (1990). Comparative fit indexes in structural models. Psychological Bulletin, 107, 238–246. Bettis, R. A., Gambardella A., Helfat, C., & Mitchell, W. (2014). Editorial: Theory in strategic management. Strategic Management Journal, 36, 1411–1413. Bernaards, C. A., & Sijtsma, K. S. (2000). Influence of imputation and EM methods of factor analysis when item nonresponse in questionnaire data is nonignorable. Multivariate Behavioral Research, 35, 321–364. Volume 20, Number 2 • April - June 2020 93

Beyer, J. M., & Trice, H. M. (1982). The utilization process: A conceptual framework and synthesis of empirical findings. Administrative Science Quarterly, 27, 591–622. Campbell, N. R. (1953). What is science? New York: Dover. Cattell, R. B. (1966). The scree test for the number of factors. Multivariate Behavioral Research, 1, 245–276. Chou, C. P., & Bentler, P. M. (1995). Estimation and tests in structural equation modeling. In R. H. Hoyle (Ed.), Structural equation modeling: Concepts, issues, and applications (pp. 37–55). Thousand Oaks, CA: Sage. Cole, S., & Cole, J. R. (1967). Scientific output and recognition: A study in the operation of the reward system in science. American Sociological Review, 32, 377–390. Colquitt, J. A., & Zapata-Phelan, C. P. (2007). Trends in theory building and theory testing: A five-decade study of the Academy of Management Journal. Academy of Management Journal, 50(6), 1281–1303. Conlon, E. (2002). Editor’s comments. Academy of Management Review, 27(4), 489–492. Corley, K. G., & Gioia, D. A. (2011). Building theory about theory building: What constitutes a theoretical contribution? Academy of Management Review, 36(1), 12–32. Davis, M. S. (1971). That’s interesting! Towards a phenomenology of sociology and a sociology of phenomenology. Philosophy of the Social Sciences, 1, 309–344. DiMaggio, P. (1995). Comments on “What theory is not.” Administrative Science Quarterly, 40, 391–397. DiMaggio, P., & Powell, W. W. (1983). The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2), 147–160. Dubin, R. (1969). Theory building. New York: Free Press. Eisenhardt, K. M. (1989a). Agency theory: An assessment and review. Academy of Management Review, 14(1), 57–74. Eisenhardt, K. M. (1989b). Building theories from case study research. Academy of Management Review, 14(4), 532–550. Gioia, D. A., & Pitre, E. (1990). Multiparadigm perspectives on theory building. Academy of Management Review, 15(4), 584–602. Greenwood, R., & Hinnings, C. R. (1996). Understanding radical organizational change: Bringing together the old and the new institutionalism. Academy of Management Review, 21(4), 1022–1054. Hallberg, N. L., & Felin, T. (2020). Untangling infinite regress and the origins of capability. Journal of Management Inquiry, 29(1), 17–32. Hambrick, D. C. (1994). 1993 Presidential address: What if the Academy actually mattered? Academy of Management Review, 19(1), 11–16. Hambrick, D. C. (2007). The field of management’s devotion to theory: Too much of a good thing? Academy of Management Journal, 50(6), 1346–1352. Haveman, H. A., Mahoney, J. T., & Mannix, E. (2019). Editors’ comments: The role of theory in management research. Academy of Management Review, 44(2), 241–243. Hu, L. T., & Bentler, P. M. (1999). Cutoff criteria for fit indexes in covariance structure analysis: Conventional criteria versus new alternatives. Structural Equation Modeling, 6, 1–55. Jensen, M. C., & Meckling, W. H. (1976). Theory of the firm: Managerial behavior, agency costs and ownership structure. Journal of Financial Economics, 3, 305–360. Judge, T. A., Cable, D. M., Colbert, A. E., & Rynes, S. L. (2007). What causes a management article to be cited: Article, author, or journal? Academy of Management Journal, 50(3), 491–506. Kaiser, H. F. (1960). The application of electronic computers to factor analysis. Educational and Psychological Measurement, 20, 141–151. Kaplan, A. (1964). Conduct of inquiry: Methodology for behavioral science. Scranton, PA: Chandler Publishing. 94 Journal of Management Research

Kilduff, M. (2006). Editor’s comments: Publishing theory. Academy of Management Review, 31(2), 252–255. Kline, R. B. (1998). Principles and practices of structural equation modeling. New York: Guilford Press. Lado, A. A., Boyd, N. G., Wright, P., & Kroll, M. (2006). Paradox and theorizing within the resource-based view. Academy of Management Review, 31(1), 115–131. MacCallum, R. C., Browne, M. W., & Sugawara, H. M. (1996). Power analysis and determination of sample size for covariance structure modeling. Psychological Methods, 1(2), 130–49. Makadok, R., Burton, R., & Barney, J. (2018). A practical guide for making theory contributions in strategic management. Strategic Management Journal, 39, 1530–1545. Marsh, H. W., & Hocevar, D. (1985). Application of confirmatory factor analysis to the study of self-concept: First- and higher-order factor models and their invariance across groups. Psychological Bulletin, 97(3), 562–582. McKinley, W., Mone, M. A., & Moon, G. (1999). Determinants and development of schools in organization theory. Academy of Management Review, 24(4), 634–648. Miller, K. D., & Tsang, E. W. K. (2010). Testing management theories: Critical realist philosophy and research methods. Strategic Management Journal, 32, 139–158. Miner, J. B. (1984). The validity and usefulness of theories in an emerging organizational science. Academy of Management Review, 9(2), 296–306. Miner, J. B. (2003). The rated importance, scientific validity, and practical usefulness of organizational behavior theories: A quantitative review. Academy of Management Learning and Education, 2(3), 250–268. Mintzberg, H. (2005). Developing theory about the development of theory. In K. G. Smith & M. A. Hitt (Eds.), Great minds in management: The process of theory development (pp. 355–372). Oxford: Oxford University Press. Ouchi, W. G. (1980). Markets, bureaucracies, and clans. Administrative Science Quarterly, 25(1), 129–141. Oxley, J. E., Rivkin, J. W., & Ryall, M. D. (2010). The Strategy Research Initiative: Recognizing and encouraging high-quality research in strategy. Strategic Organization, 8(4), 377–386. Ployhart, R. E., & Bartunek, J. M. (2019). Editors’ comments: The is nothing so theoretical as good practice – a call for phenomenal theory. Academy of Management Review, 44(3), 493–497. Poole, M. S., & Van de Ven, A. H. (1989). Using paradox to build management and organization theories. Academy of Management Review, 14(4), 562–578. Popper, K. R. (1959). The logic of scientific discovery. New York: Basic Books. Priem, R. L., & Butler, J. E. (2001). Tautology in the resource-based view and the implications of externally determined resource value: Further comments. Academy of Management Review, 26(1), 57–66. Roth, P. L. (1994). Missing data: A conceptual review for applied psychologists. Personnel Psychology, 47, 537–560. Rubin, D. B. (1996). Multiple imputation after 18+ years. Journal of the American Statistical Association, 91, 473–489. Rynes, S. L. (2005). From the editors: Taking stock and looking ahead. Academy of Management Journal, 48(1), 9–15. Sandberg, J., & Tsoukas, H. (2011). Grasping the logic of practice: Theorizing through practical rationality. Academy of Management Review, 36(2), 338–360. Shapira, Z. (2011). “I’ve got a theory paper –Do you?”: Conceptual, empirical, and theoretical contributions to knowledge in the organizational sciences. Organization Science, 22(5), 1312–1321. Smith, W. K., & Lewis, M. W. (2011). Toward a theory of paradox: A dynamic equilibrium model of organizing. Academy of Management Review, 36(2), 381–403. Suddaby, R. (2010). Editor’s comments: Construct clarity in theories of management and organization. Academy of Management Review, 35(3), 346–357. Sutton, R., & Staw, B. (1995). What theory is not. Administrative Science Quarterly, 40(3), 371–384. Volume 20, Number 2 • April - June 2020 95

Tourish, D. (2020). The triumph of nonsense in management studies. Academy of Management Learning & Education, 19(1), 99–109. Turner, J. H. (1985). In defense of positivism. Sociological Theory, 3(2), 24–30. Turner, J. H. (1989). A positivist’s reflections on the problems and prospects of general theory in sociology. International Review of Sociology, 3(2), 79–106. Venkatraman, N. (1990). Performance implications of strategic coalignment: A methodological perspective. Journal of Management Studies, 27(1), 19–41. Weick, K. (1995). Sensemaking in organizations. Thousand Oaks, CA: Sage. Wheaton, B., Muthen, B., Alwin, D. F., & Summers, G. (1977). Assessing reliability and stability in panel models. Sociological Methodology, 8(1), 84–136. Whetten, D. A. (1989). What constitutes a theoretical contribution? Academy of Management Review, 14(4), 490–495. Williamson, O. E. (1975). Markets and hierarchies: Analysis and antitrust implications. New York: Free Press. 96 Journal of Management Research

APPENDIX Survey Instrument (30 items) and Resulting Theory Evaluation Scale (25 items) After choosing a theoretical exemplar that they reported being most familiar with, respondents were given the following instructions: Please answer the following set of questions with respect to [Author(s) (year): “Title.” Publication Journal, volume(issue): page#–page#]. Please indicate the extent to which you agree or disagree with the following statements: Item # Dropped Item text N1 x The theory explored novel phenomena. N2 The theory introduced original constructs. N3 The theory proposed new relationships between constructs. N4 The theory offered different explanation for previously established relationships N5 The theory's rationale and core propositions were intriguing N6 The theory contributed to a different way of looking at organizations N7 The theory created an \"aha\" moment when proposed E1 The theory's rationale could be extended to shed light on new questions E2 The theory could be used to understand relationships at both the \"micro\" and \"macro\" levels of analysis E3 The theory as presented could be extended to a broad range of phenomena E4 x The theory as presented did not realize its full potential E5 The theory provided opportunities for further development P1 Understanding of the theory helps practitioners formulate action plans P2 A manager who understand the theory will make better decisions than one who doesn't P3 The theory is so abstract that it has limited impact on real life decisions [reverse scored] P4 The theory generates usable knowledge in the real world P5 The theory offers a good tool for business consultants P6 The theory addresses issues relevant to organizations P7 Predictions from the theory are not interesting to practitioners [reverse scored] P8 The theory leads management practice to address crucial questions F1 x The theory makes unambiguous predictions F2 The theory can be tested against facts F3 Predictions from the theory can be contrasted with empirical evidence F4 The theory's constructs are easily translated into measurable variables F5 It is easy to identify the independent and dependent variables of the theory LC1 The theory did not improve human understanding of the world [reverse scored] LC2 There were inherent contradictions in the theory [reverse scored] LC3 The propositions derived from the theory were tautological [reverse scored] LC4 x The theory's propositions were properly deduced from its assumptions LC5 x The theory provides a strong justification of its predictions Responses were recorded electronically using the following scale: Strongly Moderately Slightly Neither Slightly agree Moderately Strongly Disagree Disagree disagree Agree nor agree agree Disagree 1 2 3 56 7 4 Volume 20, Number 2 • April - June 2020 97

Endnotes 1. The co-authors divided this task so that each article was coded by a single rater. 2. We should note here the existence of a trade-off between extendibility through broader scope and the precision of a theory’s predictions (e.g., Makadok et al., 2018). We will come back to a discussion of this trade-off later on. 3. In other words, some of the ambiguity in management theorizing is unjustified. 4. The trade-offs here are with parsimony and, importantly, with logical consistency, which will also be discussed below. 5. The rise of Knowledge Management as a distinct sub-discipline in the field of management provides a case in point. 6. Note that empirical falsifiability is only one aspect of the broader concept of Popperian falsifiability (Popper, 1959). Indeed, Popper’s concept of falsifiability of a theory subsumes both internal consistency and empirical testability as defined here. 7. Some participants were members of two, or all three, of these divisions. Limiting the study to U.S.-based scholars was done to facilitate telephone follow-up during data collection. As discussed below, future studies with other samples might be warranted. 8. Respondents who reported lack of sufficient familiarity with any of the listed references were not allowed to proceed, and the survey would end for them. 9. As opposed to using regression imputation (Rubin, 1996), personal mean imputation (Bernaards & Sijtsma, 2000), mean substitution (Roth, 1994), or any other sampling distribution inference technique. 10. Note that this measurement model is also equivalent to alternative models with any four factors loading on the second-order factor, plus a covariance between the exogenous factor and the second-order factor. Any such models will produce the exact same fit as when all five factors load on the higher-order factor (e.g., Marsh & Hocevar, 1985). 11. The maximum inter-factor correlation was .67. 98 Journal of Management Research

Journal of Management Research Vol. 20, No. 2, April - June 2020, pp. 99-112 Impact of Service Failure Attributions on Dissatisfaction: Revisiting Attribution Theory Mala Srivastava and Aditi Gosain Abstract This study investigates the impact of service failure attributions on customer satisfaction in retail banking. This was done using a 2×2×2 between-subjects experimental design. The findings demonstrate that customer satisfaction after experiencing a service failure is significantly influenced by customers’ perceived attributions regarding the failure. This paper contributes to the existing literature on service failure by drawing inferences from Heider’s (1958) attribution theory. This study empirically examines how service failure attributions affect customer-service provider relationships in the Indian retail banking industry. The information presented could be used to predict the extent of customers’ post-service failure dissatisfaction. At the same time, this study shows that having an understanding of how customers form their perceptions about a service failure can help companies mitigate unfavorable customer perceptions. Keywords: S ervices Marketing, Customer Relationship Management, Service Failure, Attribution, Customer Satisfaction INTRODUCTION 1985). Because of these fundamental characteristics, Increased competition and worsening economic banks often fail to meet their customers’ and business conditions have forced organizations expectations (Boshoff, 1997), thus resulting in to develop new products and services to increase a relatively higher frequency of dissatisfaction revenue. Organizations must also constantly revisit amongst customers (Shostack, 1977). As a result, and redesign their customer-retention strategies. service failure perceptions are frequent. A service Inexorably, companies occasionally fail to meet failure is a service incident that leads customers customer expectations when delivering their to perceive that the service has failed to meet their services. Numerous studies point out that service expectations. failures cause customer complaints and negatively Marketers must take customer complaints and affect customer retention and acquisition. service failures seriously. Service failures can lead to declines in customer confidence, increases in The inseparability and heterogeneity characteristics customer defection, and negative word-of-mouth. of the services provided by banks make the They also increase costs, as the unsatisfactory services management of customer relationships especially have to be re-performed (Parasuraman et al., 1991). difficult in the retail banking sector (Zeithaml et al., Service failures also have internal effects, such as perceptions of role ambiguity, perceptions of lack Mala Srivastava of empowerment, and frustration in dealing with IIM Kashipur, Kashipur, District Udham Singh Nagar dissatisfied or angry customers (Yavas et al., 2003). Kashipur, Uttarakhand – 244713 India However, the relationship between a service failure Aditi Gosain and customer’s response to the failure partially Sinhgad Institute Of Business Administration And depends on the customer’s perception about what Research (Sibar) caused the service failure incident. Considering this, Kondhwa, Pune, Maharashtra – 411048 India

the present study draws inferences from Heider’s inferences on the customer’s post-failure behavior. (1958) attribution theory to study how customers’ The information presented can help marketers perceived causal inferences regarding a service failure develop effective recovery strategies. affect their satisfaction with the service provider. Attribution is the process of assigning causes to CONCEPTUAL FRAMEWORK AND an incident (Weiner, 1992). It is fundamental HYPOTHESES to human thought and is one of the most widely studied phenomena in social psychology (Lawrence Services marketing has matured as a distinguished & Nohria, 2001; Main et al., 2007; Madrigal, field of research since the publication of Shostack’s 2008). Attribution theory views people as rational (1977) article “Breaking Free from Product information processors whose actions and behaviors Marketing,” which highlighted the challenges of are influenced by their causal inferences (i.e., causes services marketing. The early services marketing attributable to a particular event or situation) literature focused on identifying the intangibility (Weiner, 1980). Thus, a customer’s response to and other unique characteristics of services, namely failure is not based merely on the fact that the inseparability, heterogeneity, and perishability product or service has failed. The customer will also (Zeithaml et al., 1985). These characteristics are try to determine why the failure has occurred when considered the basic factors that differentiate between responding to it (Folkes, 1984). products and services. Each of these characteristics Many researchers have studied attribution theory poses specific problems in the marketing of services. from a multidimensional perspective and in relation Also, due to these inherent characteristics of services, to satisfaction and other behavioral dimensions they are more prone to failure than products. Service such as word-of-mouth intentions, repatronage failure has been defined as an observable decline intentions, and complaining behavior (Richens, in quality in what has been termed as the service 1983; Curren & Folkes, 1987; Oliver & DeSarbo, encounter (Smith, 1997). 1988; Smith et al., 1999). However, there are some limitations in the extant Service Failure literature regarding the impact of customers’ A service failure occurs when the quality of a service attributions on their responses. For one, varied is worse than expected by the customer (Palmer et conceptualizations of attributions have been formed al., 2000; Hess, et al., 2003). Service failures can by selecting either just one or two of its dimensions. cause customer dissatisfaction (Parasuraman et al., Furthermore, some researchers have looked at only 1985). They can also prompt customers to speak one level of attribution dimensions while ignoring negatively about the service provider (Richins, others (e.g., stable vs. unstable). As a result, it is not 1983) or to defect to alternative service providers known which of these dimensions has the greatest (McCollough et al., 2000). impact on customer responses. Also, most previous In today’s competitive world, service firms cannot studies have considered only the main effects of afford to ignore service failures. This is primarily attribution dimensions on the dependent variables. because service failures are (i) prevalent, (ii) However, some attribution dimensions may have a memorable, and (iii) lead to customer defection multiplicative effect on the dependent variables. (Smith, 1997). Moreover, service failures can impair To fill these gaps in the literature, the present a firm’s profitability and reduce customers’ loyalty study explores the interaction effects of different towards the firm. Thus, firms must understand the attribution dimensions. The present study also mental processes that impact customers’ reactions examines the causal inferences that a customer to service failures. Customer’s understanding of applies to service failures and the impact of such the causes of a service failure often determines the customer’s attitude towards the service provider. 100 Journal of Management Research

Customers’ post-service failure attributions can opposed to attributions or the satisfaction itself ) as either intensify or mitigate negative consumer a dependent variable. responses and affect overall customer satisfaction (e.g., Curren & Folkes, 1987; Smith & Bolton, Attribution Theory 1998; Smith et al., 1999; Hess et al., 2003). Attribution theory can be used to understand how customers’ perceived reasons for a product failure Customer Satisfaction influence their responses to the failure (Folkes, Customer satisfaction is a customer’s subjective 1984). The theory is concerned with all aspects assessment of a product or service after using of causal inferences: how people arrive at causal the product or experiencing the service inferences, what sort of inferences they make, and (Westbrook, 1980). Theoretically, satisfaction is the consequences of these inferences (Folkes, 1984, a post-consumption outcome. According to the 1988). The seminal concepts underlying attribution disconfirmation of expectations model (Oliver, theory can be found in Heider’s (1958) book “The 1980), satisfaction is usually based on a consumer’s Psychology of Interpersonal Relations.” comparison between their pre-purchase expectations Heider’s analysis of different types of causes of and the observed performance of a product or failure has been extended by other researchers service. The disconfirmation of expectations model (Weiner et al., 1985). These researchers have is widely applied and has received much empirical determined that occurrence of an event gives rise support (Churchill & Surprenant, 1982; Oliver & to a search for the causes of the event. Following DeSarbo, 1988; Westbrook & Oliver, 1991). an event, a cognition (attribution)-emotion- The literature presents two broad conceptualizations action temporal sequence occurs, by which causal of satisfaction: transaction satisfaction and ascriptions produce an affect. This search for cumulative satisfaction. Transaction satisfaction causation (attribution) has been studied in terms refers to a customer’s evaluation of a specific of three dimensions (Weiner et al., 1979; Weiner, experience (e.g., Anderson & Fornell, 1994; Oliver, 1980) listed below: 1981; Parasuraman et al., 1994). Cumulative Locus: Whether the failures are perceived to be satisfaction denotes a consumer’s satisfaction caused by the customer (internal locus) or the with their relationship with a service provider. service provider (external locus) Cumulative satisfaction is built over time as a Stability: Whether the causes are perceived to be customer experiences multiple instances of services permanent (stable) or temporary and infrequent from a service provider (e.g., Bitner & Hubbert, (unstable) 1994; Bolton & Drew, 1991; Crosby & Stephens, Controllability: Whether the causes are perceived 1987; Oliva et al., 1992). to be preventable (controllable) or unpreventable The relationship between multiple service instances (uncontrollable) and customer satisfaction has been explored in some It is expected that the different types of attributions notable research studies (Bitner, 1990; Bolton & can have different effects on a customer’s post- Drew, 1991; Oliver & DeSarbo, 1988; Oliver & failure response. The present study proposes that the Swan, 1989). Our view is that customers continue three dimensions of causality (i.e., locus, stability, to revise their cumulative satisfaction levels towards and controllability) individually affect post-service service providers as they experience satisfactory failure dissatisfaction. The research hypotheses services, service failures, and recoveries. Based on have been framed as expressed in the following what is already known about consumer complaining behavior, this study uses ‘changes’ in satisfaction (as Volume 20, Number 2 • April - June 2020 101

subsections so that their impacts can be studied H4: Stable attributions have a greater impact on individually. post-failure dissatisfaction than unstable attributions. Impact of Locus Attributions on Impact of Controllability Attributions on Post-service Failure Dissatisfaction Post-service Failure Dissatisfaction The locus dimension of attributions determines The controllability dimension of attributions refers whether customers blame themselves or the service to customers perceptions regarding the level of provider for a service failure. Generally, customers control that the service provider had to prevent a find it difficult to blame themselves for service failure from occurring (Weiner, 1985; 2000; Hui & failures (Hui & Toffoli, 2002). Ye (2005) proposed Toffoli, 2006). Customers who perceive that a service that customers tend to give themselves credit for failure could have been prevented by the provider good outcomes and ascribe blame to external are expected to be more dissatisfied than customers factors for poor outcomes. Thus, it is expected that who think that little could have been done to avoid customers’ reactions to service failures depend on the failure (Folkes, 1984; Hess et al., 2003; del Río- whether they find themselves (internal locus) or the Lanza et. al., 2009). Consequently, we propose the service provider (external locus) at fault. following hypotheses: H1: Locus attributions impact customers’ post- H5: Controllability attributions impact customers’ failure dissatisfaction. post-failure dissatisfaction. H2: External locus attributions have a greater H6: Controllable attributions have a greater impact impact on post-failure dissatisfaction than internal on post-failure dissatisfaction than uncontrollable locus attributions. attributions. Impact of Stability Attributions on Interaction Effects Post-service Failure Dissatisfaction All three causal attributions are expected to directly The stability dimension of attributions has to do affect satisfaction. However, satisfaction can vary with the customer’s perception of the circumstances based on the type of attributions made. Also, the that led to the failure as either infrequent or attribution dimensions might have interaction persistent. The principle of consistency is a strong effects on satisfaction. Although interaction effects contributor to the formation of attributions are commonly examined in social psychological (Heider, 1958). For example, if a customer’s studies, such effects have been considered in very previous experiences with a service provider have few studies in the customer satisfaction literature been satisfactory, a sudden drop in quality will be (Folkes, 1984, Oliver & DeSarbo, 1988). Thus, considered as having an unstable cause (Weiner, the present study explores the interaction effects of 2000). In such cases, the customer is expected locus, stability, and controllability on post-failure to forgive the service provider and not be highly dissatisfaction. dissatisfied. However, if customers feel that the H7: Locus and stability have an interaction effect service failure is likely to persist, they will probably on post-failure dissatisfaction. be highly dissatisfied with the service. Thus, we H8: Locus and controllability have an interaction hypothesize the following: effect on post-failure dissatisfaction. H3: Stability attributions impact customers’ post- H9: Stability and controllability have an interaction failure dissatisfaction. effect on post-failure dissatisfaction. 102 Journal of Management Research

H10: Locus, stability, and controllability have an as the difference between participants’ pre- interaction effect on post-failure dissatisfaction. service satisfaction levels and post-service failure satisfaction levels. EXPERIMENTAL PROCEDURE A 2×2×2 between-subjects pretest-posttest design Satisfaction was used. Role-playing methodology was Consumer satisfaction was measured in two stages: employed by using hypothetical scenarios. pre-service failure (PSAT) and post-service failure Each attribution dimension was manipulated (FSAT) because pre- and post-service failure are at two levels (i.e., locus: internal or external, considered two distinct experiences. Measures for stability: stable or unstable, and controllability: PSAT and FSAT were based on the measures used controllable or uncontrollable) (Appendix 1). in previous studies (Bitner & Hubbert, 1994; Oliver Each scenario contained a unique combination & Swan, 1989; Smith, 1997). PSAT and FSAT were of service failure attributions. For example, one measured using a 4-item bipolar adjective scale. The of the scenarios was stable, uncontrollable, and dependent variable in the study was the difference involved an internal locus. The scenarios covered between the PSAT and FSAT (i.e., the change all possible combinations of attributions (Table in satisfaction level after the service failure). This 1). The scenarios were randomly assigned to the variable was labeled as DSAT. A high DSAT value participants, with one scenario given to each indicated a high level of dissatisfaction (Appendix 2). respondent. Manipulation of Independent Factors Sampling The three service failure attributions (i.e., locus, The respondents of this study were retail banking stability, and controllability) were treated as services customers. The respondents were carefully independent factors in the construction of the selected using predefined subject selection criteria service failure scenarios for this experiment. These (e.g., having sufficient previous experience with independent factors were manipulated through retail banking services) depending on the nature of scenario descriptions (For instance, scenario 1 the scenarios generated through a pilot study. expressed that the service failure was of an unstable An adequate sample size was identified to nature through inclusion of the statement, “You ensure that a statistically sufficient number of had never faced such a problem earlier.”). All three respondents were obtained for each scenario type. dimensions were manipulated at two levels each, Approximately 50 respondents were required for thus creating a 2×2×2 factorial design. each of the eight scenarios, meaning that at least The dependent variable was the level of change 400 respondents were needed for the study. Our in dissatisfaction. This variable was measured selection process allowed us to choose respondents Table 1: Combinations of Attributions for Service Failure Scenarios Internal Locus (L0) External Locus (L1) Controllable Uncontrollable Controllable Uncontrollable Failure (C0) Failure (C1) Failure (C0) Failure (C1) Stable (S0) L0C0S0 L0C1S0 L1C0S0 L1C1S0 Unstable (S1) L0C0S1 L0C1S1 L1C0S1 L1C1S1 Volume 20, Number 2 • April - June 2020 103

who were able to satisfactorily play the role of the attributions on customer satisfaction. It was customer as depicted in the scenario assigned to assessed whether the experimental stimuli brought them. Some important criteria used for this purpose about any changes in customers’ post-stimuli were as follows: satisfaction and behavioral intentions. Each • The participants should have a reasonable dimension of attribution was manipulated at two levels. Therefore, for each of the attribution level of experience with Indian retail banking dimensions, an investigation was conducted to services. assess whether post-failure customer satisfaction • Participants should be familiar with the service changed at different levels of manipulation. This failure scenarios and should find it easy to relate analysis revealed the relative effects of each level of to them. manipulation on change in customer satisfaction. DATA COLLECTION AND ANALYSIS ANALYSIS AND RESULTS Data collection included random assignment of A univariate analysis was conducted to test experimental stimuli. Explicit instructions were differences between attribution levels and to study given to the respondents to ensure that internal their interaction effects. validity requirements were met. The questionnaire included an introduction, a request letter, and a Effects of Locus on Post-Service Failure set of general instructions for the participants to Dissatisfaction follow. Aside from these sections, the questionnaire Analysis showed that the main effect of locus comprised of four main parts: attributions on dissatisfaction was statistically Part A: This section dealt with descriptive measures, significant (F = 115.24, p = 0.001), thus confirming such as the age, gender, and income of respondents. H1. The changes in DSAT levels among the These questions were optional. The responses external locus respondents and internal locus were analyzed only at aggregate levels so that respondents were 1.626 and 0.630, respectively. confidentiality was ensured regarding respondents’ The tested t and p values were 7.484 and 0.001, personal information. respectively, which indicated a significant difference Part B (Pretest): This section contained a pre-service in post-failure satisfaction scores between the failure evaluation to measure the respondents’ respondents with internal locus attributions and satisfaction with the service provider and behavioral those with external locus attributions. This lends intentions before the failure occurred. support to H2, which proposes that external locus Part C (Posttest): This section included a attributions have a greater impact than internal post-service failure evaluation to measure the locus attributions on post-failure dissatisfaction. respondents’ satisfaction with the service provider and behavioral intentions after the failure occurred. Effects of Stability on Post-Service Failure Part D: This section contained a believability and Dissatisfaction realism check. The purpose of this section was to The analysis validated H3, as it indicated that assess how realistic the service failure scenarios stability attributions had a significant impact on seemed to the respondents. This section also satisfaction (F = 43.92, p = 0.001). The mean assessed how easy participants found it to adapt to DSAT scores for stable scenario respondents the role of the customer depicted in the scenario and unstable scenario respondents were 1.435 assigned to them. and 0.820, respectively. The tested t and p values A 2×2×2 factorial ANOVA was used to investigate were 4.286 and 0.001, respectively, indicating a the impacts of locus, stability, and controllability significant difference in post-failure satisfaction 104 Journal of Management Research

scores between respondents with stable attributions and 0.994, respectively. The tested t and p values and those with unstable attributions. This lends were 1.601 and 0.070, respectively, which indicates support to H4, which proposes that stable insignificant difference in post-failure satisfaction attributions have a greater impact than unstable scores between the respondents with controllable attributions on post-failure dissatisfaction. attributions and respondents with uncontrollable Effects of Controllability on Post-Service Failure attributions. Thus, H6 was not supported. Dissatisfaction The main effect of controllability attributions on Interaction Effect satisfaction was significant (F = 8.355, p = 0.004), Of the four hypotheses related to interaction thus supporting H5. Our related hypothesis (H6) effects (i.e., H7, H8, H9, and H10), only H7 was was that controllable attributions would have a supported. This study employed a 2 × 2 × 2 factorial greater impact than uncontrollable attributions ANOVA to test the extent to which different on post-failure dissatisfaction. The mean DSAT attribution types affect customer satisfaction. As scores of controllable attribution respondents and shown in Table 2, the interaction effect of locus and uncontrollable attribution respondents were 1.262 stability is statistically significant (F = 9.237, p = Table 2: Univariate Analysis (Tests of Between-Subjects Effects) Source Type III Sum of Squares df Mean Square F Sig. 22.957 .001 Corrected Model 177.959 7 25.423 590.469 0.001 115.243 0.001 Intercept 653.884 1 653.884 43.92 0.001 0.004 LOCUS 127.621 1 127.621 8.355 0.002 9.237 0.476 STABLE 48.636 1 48.636 0.508 0.624 0.241 0.723 CONTROL 9.252 1 9.252 0.125   LOCUS * STABLE 10.229 1 10.229       LOCUS * CONTROL 0.562 1 0.562 STABLE * CONTROL 0.267 1 0.267 LOCUS * STABLE * CONTROL 0.139 1 0.139 Error 589.136 532 1.107 Total 1370.875 540   Dependent Variable: DSAT Table 3: Parameter Estimates Parameter Mean Sd B Std. Error t Sig. Intercept 2.236 .146 15.319 .001 [LOCUS=0] Internal .6653 .99 1.378 .184 .001 [LOCUS=1] External 1.590 1.23 7.284 [UNSTABLE=0] 1.12 .885 .206 .001 [STABLE=1] .8036 1.20 4.286 [UNCONTROL=0] 1.59 1.10 .201 .152 .073 [CONTROL=1] 1.27 1.601 .967 1.15 Volume 20, Number 2 • April - June 2020 105

0.002). However, the interaction effects of locus has a significant effect on satisfaction in a and controllability (F = 0.508, p = 0.476); control disconfirmation situation. They also reported that and stability (F = 0.241, p = 0.624); and locus, customers are less satisfied with a service when the stability, and controllability (F = 0.125, p = 0.723) failure was controllable than when the failure was were insignificant. Thus, H8, H9, and H10 were uncontrollable. However, in the present study, an not supported. insignificant difference in satisfaction was observed between customers with controllable attributions DISCUSSION and those with uncontrollable attributions. Main Effects Our findings are also contrary to those of Hamilton The results supported H1, H2, H3, H4, and H5. (1980). Hamilton’s study showed that customers’ H6 was not supported. The results indicated that behavior is affected by their perceptions of whether customers’ dissatisfaction levels change according the service organization could have prevented a to their service failure attributions. Most of the failure. Hamilton suggested that a customer is differences observed occurred in the expected likely to be more forgiving if the failure was due to directions. The findings suggest that attribution an uncontrollable situation than if the failure was principles are useful for understanding customers’ preventable. post-service failure dissatisfaction. The support for Perhaps the findings of the current study occurred H1, H3, and H5 provided in the present study because the controllability attribution’s effect on corroborate the findings of Iglesias (2009), who customer satisfaction depends on the nature of reported that locus, stability, and controllability the service being provided. The banking industry attributions affect satisfaction after a service delivery is a highly regulated and process-driven industry. failure. As such, bank customers might believe that banks Similarly, our results regarding H2 and H4 are have a high level of control over their operations in line with the results of Bitner’s (1990) study, and, therefore, are unlikely to face uncontrollable which demonstrated that a customer displays situations. a less negative response to service failure when they attribute the failure to (i) internal rather Interaction Effects than external causes, (ii) uncontrollable rather H7 posited that there is a significant interaction than controllable factors, and (iii) unstable rather between locus and stability attributions, and than stable conditions. Additionally, the findings this hypothesis was supported by our results. reported in the present study are in agreement with This finding is in line with the work of Tsiros the findings of several researchers who have found et al. (2004), who studied the interaction that external attribution leads to higher levels of between stability and responsibility (in their dissatisfaction than internal attribution (Krishnan study, responsibility incorporated locus and & Valle, 1979; DeCarlo & Leigh, 1996; Valenzuela controllability). They found a significant interaction et al., 2005; Grewal et al., 2008). Our support effect between these variables. Oliver and DeSarbo of H4 upholds the work of Meyer (1980), who (1988) also studied the interaction between stability proposed that customers who identify the source and locus attributions. However, they found that of a failure as stable experience a stronger sense the interaction effect between these two variables of dissatisfaction than customers who identify the was non-significant. source of a failure as unstable. The other interaction effects (H8, H9, and H10) Our results regarding H5 did not agree with investigated in the present study were insignificant. the results of a study conducted by Walton and This is probably because of the confounding effect Hume (2012). They reported that controllability of the controllability attribution. 106 Journal of Management Research

IMPLICATIONS AND LIMITATIONS ‘true test of commitment’ and their ability to This study examined how a customer’s perceived manage difficult situations. attributions of a service failure affect the This study provides some distinctive and relevant relationship between the customer and service inferences for decision-makers in the Indian retail provider. This research helps us identify which banking industry. Namely, our results demonstrate service failure attribution dimensions (and which that the structure of a customer’s attributional interactions between them) have the greatest impact beliefs will likely govern their response to a on customer satisfaction. Making this kind of service failure. Therefore, organizations need to identification can be helpful in managing the types manage how customers form these types of of service failures that have the most adverse effects beliefs during and after a service failure. This can on the relationship between the customer and the be done by structuring the flow of information service provider. in a timely, precise, and accurate manner. This If service failure attributions are managed will sway customers to form less-negative post- correctly, they may result in higher customer failure perceptions towards the service provider. A satisfaction and loyalty than if the failure had noteworthy limitation of the present study is the not occurred. This is because while assessing a limited number of variables. The impact of service service failure, a customer can assess the service criticality and severity, although important, are orientation of the service provider and can learn difficult to assess in an experimental study. It would about their processes and policies. Thus, the be difficult to exclude customers with individual customer’s trust in the service provider could exit barriers and high search costs. increase through an appreciation of the provider’s Volume 20, Number 2 • April - June 2020 107

APPENDIX 1 Service Failure Scenarios Scenario 1 Locus (L): External (L1), Stability (S): Stable (S1), Controllability (C): Controllable (C1) You had placed a request on Internet Banking for the creation and delivery of an urgent Demand Draft to your residential address. You find that the demand draft has not been delivered in the stipulated time as mentioned on the website of your bank, though the required amount has been debited from your account for the Demand Draft. You had never faced such a problem earlier. You believe that the bank should be blamed for this delay, you also feel that the bank could have taken enough care to avoid this delay. Scenario 2 Locus (L): External (L1), Stability (S): Unstable (S0), Controllability (C): Controllable (C1) You had placed a request on Internet Banking for the creation and delivery of an urgent Demand Draft to your residential address. You find that the demand draft has not been delivered in the stipulated time as mentioned on the website of your bank, though the required amount has been debited from your account for the Demand Draft. You had faced such a problem at least three times earlier also. You believe that the bank should be blamed for this delay; you also feel that the bank could have taken enough care to avoid this delay. Scenario 3 Locus (L): External (L1), Stability (S): Stable (S1), Controllability (C): Uncontrollable (C0) You had deposited a high value cheque in your bank branch, to be credited in your account. After the stipulated time for cheque clearing, you realize that that the cheque amount has not been credited into your account. On making the enquiry you are informed that although there was no problem with the cheque, there were some technical difficulties in the inward clearing processes due to which your cheque was not processed. You have faced similar types of problems in this bank at least three times. You are aware that such errors are beyond the control of the bank branch. Scenario 4 Locus (L): External (L1), Stability (S): Unstable (S0), Controllability (C): Uncontrollable (C0) You had deposited a high value cheque in your bank branch, to be credited in your account. After the stipulated time for cheque clearing, you realize that that the cheque amount has not been credited into your account. On making the enquiry you are informed that although there was no problem with the cheque, there were some technical difficulties in the inward clearing processes due to which your cheque was not processed. You 108 Journal of Management Research