Process and Variance Modeling

Abstract

Models are an important component of research design that serve as intermediaries between theories and data, often directing decisions about methods and statistics. This article discusses the basic differences and assumptions associated with process and variance models as a way of introducing the four articles contained within this special issue of Family Business Review on “Process and Variance Methods.” Specifically, we highlight three key issues regarding modeling—time and causality, measurement and operationalization, and model specification—making specific ties to the challenges often associated with family business research.

Keywords

family business modeling process research methods variance

Introduction

Family business, as a field of study, has undergone significant growth over the past several decades (Short, Sharma, Lumpkin, & Pearson, 2016). Growth can be seen in terms of the total number of studies published, the impact of these studies on the academic community, and the number of conferences and journals dedicated to family business. For instance, within the business and management categories of Thomson Reuters’s Web of Science Core Collection, the topic of “family business” is associated with 221 articles published in 2015, which is up from only 15 in 1996. Also, the number of citations associated with these articles has increased from a meager 37 in 1996 to 5,190 in 2015.

With such growth comes evolution toward more complex research questions and theories, which then requires the development and utilization of new and better research methods and analytics. Indeed, to continue to advance and legitimize the field, the family business research community needs to incorporate more sophisticated and novel methodological approaches to address fundamental questions about family businesses, as well as emerging ones (Evert, Martin, McLeod, & Payne, 2016). As such, the goals of this special issue on “Process and Variance Methods in Family Business Research” are to (1) inform researchers about key novel and advanced methods and techniques that may prove particularly useful in their study of family businesses and (2) challenge scholars to seek out, develop, and use these types of enhanced methodological practices to increase the confidence in and relevance of research findings.

In an effort to help achieve these goals and familiarize readers with the articles contained within this special issue, this introductory article focuses on models that guide methodological and empirical decisions. Essentially, models serve as intermediaries or mediators between theories and data, often driving decisions about methods and empirics (Morrison & Morgan, 1999; Van de Ven, 2007). For our purposes, models serve as instruments or tools through which we empirically investigate, and subsequently learn about, family businesses.

There are two basic types of models—termed process and variance models—that are used in research. Each type represents a different perspective regarding research questions that are being asked about a particular family business issue or phenomenon. Process models examine events and narratives to address the question, “How does the issue or phenomenon change over time?” Variance models, on the other hand, examine the relationships between independent and dependent variables to address the question, “What are the antecedents and consequences associated with the issue or phenomenon?” Both questions are important, incommensurable, and complementary (Van de Ven, 2007). Therefore, both the “how” and the “what” questions, along with their associated models, pose unique methodological challenges for family business researchers, but they cannot be truly divorced from the other.

This introductory article briefly reviews the unique issues and assumptions surrounding process and variance models, highlighting the need for family business scholars to carefully consider these differences when making methodological and empirical decisions. Specifically, we discuss three broad considerations that should be deliberated when making modeling decisions. As part of this discussion, we introduce the four articles contained within this special issue by summarizing them and linking them to the challenges associated with modeling research questions within the family business domain. These four articles are the result of a rigorous review process. First, the special issue editorial team selected 10 of the most promising proposals submitted in response to the call for papers and solicited full-length papers. Then, the papers were sent through the double-blind peer review process, where both family business and research methods experts were used to provide critical and developmental feedback. In all cases, the special issue editorial team made decisions collectively.

Process and Variance Models

Process and variance models are different largely because they contrast in their fundamental epistemological approach. Process models seek to explain the sequencing of events that lead to some outcome. Research questions driving process models ask how and why relevant constructs change over time; this can be based on established theoretical perspectives or lead to new ones. Indeed, exploring how and why using a process model can lead to new theory development, when prior theory does not yet exist. Variance models, on the other hand, seek to explore and explain relationships between independent and dependent variables, asking research questions about what the antecedents and consequences of some key construct may be. Typically, variance models theorize about the nature of relationships by drawing on one or more established frameworks; theoretical contributions often are more incremental when using variance models (Van de Ven, 2007). See Table 1 for a summary of the basic differences between process and variance models.

Table 1.

Basic Differences Between Process and Variance Models.

Comparison point	Process model	Variance model
Research question	How does an issue emerge or change over time?	What are the antecedents and/or consequences of an issue?
Epistemology	Event-driven and narrative	Outcome-driven and logico-scientific
Basic purpose	Explain a sequence of events based on a narrative or story	Explain changes in dependent variables in terms of causal independent variables
Unit of analysis	Events within which entities (i.e., individuals or collectives) participate; entities may change as a result of events	Entities (i.e., individuals or collectives) that possess specific variable attributes
Temporality	Critical for understanding linkages between events and patterns	Largely immaterial among independent variables
Causality	Can involve both immediate and more distal causation; explanations can be complex and layered in different time scales	Emphasizes immediate causation, which is indicated by covariation, temporal precedence, and no spurious factors

To exemplify these two types of models, consider the key family business topic of succession, which can be considered from both a process and variance perspective. Defined as “the actions and events that lead to the transition of leadership from one family member to another in family firms” (Sharma, Chrisman, Pablo, & Chua, 2001, p. 21), a process model might examine succession dynamics as they occur within, between, or across the phases, which are commonly labeled as (1) ground rules and first steps, (2) nurturing and development of successor, and (3) handoff or transition (Daspit, Holt, Chrisman, & Long, 2016; Le Breton-Miller, Miller, & Steier, 2004). For instance, in a recent qualitative study of 19 family businesses, Cater, Kidwell, and Camp (2016) examine how social dynamics in successor teams, which include multiple possible successors, can take either a positive or negative track that respectively results in either enhanced organizational commitment or dissolution of the team and, potentially, the firm.

A variance model, conversely, would examine the factors that may result in a higher likelihood of a key succession outcome (e.g., intention toward succession, formal succession plan, successful succession); factors in a variance model might include incumbent and/or successor attributes (e.g., personality, motivation, education), family characteristics (e.g., climate, identity, harmony), or aspects of the business and industry (e.g., performance, age, dynamics). For example, De Massis, Sieger, Chua, and Vismara (2016) use hierarchical regression analysis to examine if situational (e.g., number of children, number of family shareholders) and individual (e.g., incumbent emotional attachment) factors are related to incumbents’ attitude toward intrafamily succession.

While fundamentally different, process and variance models are complementary and are both important for the furtherance of family business research. Indeed, scholars often use—implicitly or explicitly—theoretical arguments that describe a process when developing hypotheses regarding the variance relationship between two variables. Likewise, process studies tend to uncover new questions or constructs that need further examination through variance methods. Despite (or perhaps because of) their complementary nature, process and variance models are often confused with one another, or mixed together in inappropriate ways (Van de Ven, 2007). For instance, one might pose the question, “How does a family firm develop a succession plan?” but then examine various constructs that are antecedents of the existence of a succession plan using a cross-sectional retrospective survey design. This example demonstrates that there is some disconnect between the research question asking “how” and the research design that analyzes “what.” Ultimately, a misfit between the research question/theory and the employed methodological design will result in misspecified or erroneous conclusions, which can have negative implications for the individual study, such as reducing the chance of the study being published (Bono & McNamara, 2011). More important, however, poorly designed and executed research will fail to meaningfully and cohesively contribute to the family business field as a whole, which can unintentionally influence family business practices in negative ways.

Key Modeling Considerations

While understanding the basic differences between process and variance models is important for proper research design, there are many related issues that should be considered when making research design decisions. Although a full discussion of these issues is beyond the scope of this article, we maintain that three key considerations are especially relevant to family business research.¹ Specifically, the three research design considerations we address include (1) time and causality, (2) measurement and operationalization, and (3) model specification and data. These modeling considerations serve as a platform for our subsequent summaries of the articles included in this special issue on process and variance methods.

Time and Causality

While different, both process and variance models incorporate time in research designs, although not always explicitly. Furthermore, most studies address some level of change, which assumes or implies causality. For process models, the temporal sequence of events is an essential component of the design. Consideration must be given to temporal linkages between events and the potential patterns of activities that develop. In practice, these patterns often reflect a developmental or emergent perspective. Furthermore, the unit of analysis—be it an entity, attribute, or event—is not assumed to remain static over time, because the meaning of any factor or event examined may change. For example, Colli (2012) discusses the complexities associated with family firm performance, where the meaning of performance may change across time (e.g., from survival to growth to reputation), particularly as new generations of family members take control of the business. This perspective differs from variance models, which generally suggest that one static factor causes another, and these relationships are presented in propositional or hypothetical “if-then” statements. For variance models, causality is indicated by covariation, temporal lags between variables, and the absence of spurious factors (Van de Ven, 2007).

Perhaps the most important implication of considering time and causality in modeling decisions concerns the characteristics of the data itself. For both process and variance models, longitudinal data are typically required. While there may be variance modeling cases where cross-sectional data may be appropriate, “researchers simply cannot develop strong causal attributions with cross-sectional data, nor can they establish change, regardless of which analytical tools they use” (Bono & McNamara, 2011, p. 657). As such, careful consideration should be given to data appropriateness and availability prior to moving forward with any research project. Additionally, the availability of appropriate data analytic techniques should be considered. In other words, researchers must ask if there are readily available and generally accepted data collection and analyses techniques that can address the research question being posed. If not, it might be prudent to change the nature of the research question.

This challenge of longitudinal data collection and analytical technique availability is directly addressed by Anglin, Reid, Short, Zachary, and Rutherford (2017) in their article titled “An Archival Approach to Measuring Family Influence: An Organizational Identity Perspective,” which is published in this issue. Specifically, these scholars demonstrate how archival sources of data (e.g., shareholder letters of publicly traded companies) can be analyzed using content analysis techniques to provide a longitudinal examination of constructs that heretofore have not been available to family business scholars. In this study, measures for family visibility, transgenerational sustainability, and family self-enhancement were developed using an organizational identity perspective and validated. Furthermore, Anglin et al. (2017) used random coefficient modeling (RCM) to examine variation in these measures across time. RCM is an underutilized technique that can be very useful because it allows researchers to examine variance at multiple levels, including time. Specifically, RCM allows for the examination of nested data (e.g., families within firms, firms within industries) as well as changes in variables across time (Bliese & Ployhart, 2002). As such, RCM techniques allow scholars to address concerns of time and causality and may be used to test research questions using either process or variance modeling. Indeed, such techniques are particularly important for family business research because of the omnipresence or salience of temporal issues in family firms (Brigham, Lumpkin, Payne & Zachary, 2014; Sharma, Salvato, & Reay, 2014).

Measurement and Operationalization

When designing research models, numerous decisions should also be considered with regard to measurement and operationalization. Determining definitions, boundary conditions of key constructs, and methods for measuring those constructs are key challenges for any empirical research project. For methods specifically, sampling characteristics (e.g., student sample, family member vs. nonfamily member informant), decisions regarding the utilization of new or existing measures, and data-gathering procedures (e.g., survey vs. experimental) should be carefully considered prior to the data collection and analysis. Oftentimes, significant problems raised in the review process could have been avoided if proper consideration had been given to construct measurement and operationalization when developing the research model.

Perhaps most important for process models are definitions and boundaries. Even the very nature of the term process can be confusing. Process can refer to either a “category of concepts of individual and organizational actions” or as “a sequence of events or activities that describe how things change over time” (Van de Ven, 2007, pp. 196-197). Often, the first definition is used to describe variables—if and to what extent change occurs over time—that are a part of a larger organizational process. For instance, we may discuss the succession planning process, but rather than focus on how and in what order the process occurs, we examine if some variables contained within the larger succession process change. From this type of perspective, scholars may deconstruct the nature of a larger sequence of events by examining the causes and consequences of singular events in the overall sequence. However, to truly understand the how associated with any process, the researcher must take a historical or narrative perspective, systematically examining the progression of activities or events over time.

For variance models, measurement concerns regarding reliability and validity are particularly important. All four articles included in this special issue make a contribution in this area. For example, Anglin et al. (2017) build on the family influence framework of Zellweger, Nason, Nordqvist, and Brush (2013), which highlights how influence is reflected in organizational identity without reliance on individual self-reports and their related biases. Self-report bias, along with the closely associated common method bias, are problematic because they threaten the validity of research, which may hinder theory development and progress as a research field. Of particular concern for family business scholars is the tendency for individuals to respond to questions—in person (e.g., interviews) or not (e.g., surveys)—in socially desirable ways (cf. Moorman & Podsakoff, 1992) leading to under- or overinflated responses and inaccurate conclusions.

The second article within this special issue also contributes to the literature through the development of validated measure. Neubaum, Thomas, Dibrell and Craig’s article (2017) titled, “Stewardship Climate Scale: An Assessment of Reliability and Validity,” develops and validates a new measure termed Stewardship Climate, which is defined as the extent to which individual employees perceive their firm’s policies, practices, and procedures foster stewardship behavior (e.g., other-focused and prosocial) and stewardship values (e.g., altruism; Neubaum et al., 2017). In developing this six-factor, 18-item scale, the authors theoretically ground their work in both stewardship theory (Davis, Schoorman, & Donaldson, 1997) and organizational climate research (Schneider, 1975) to address the perceptions of steward-like behaviors across the organization. Stewardship climate is conceived as a second order construct that is composed of six dimensions: intrinsic motivation, organizational identification, use of personal forms of power, collectivism, low power distance, and involvement orientation.

Stewardship climate can exist in any type of firm—both family and nonfamily. Likewise, it can be minimal or absent in either setting. Yet family firm research often claims stewardship behaviors as a potential differentiator in family firm behaviors, goals, strategies, and outcomes (Madison, Holt, Kellermanns, & Ranft, 2015). In essence, family firm researchers have assumed that families and family firms may act in a more altruistic and steward-like manner than nonfamily firms. However, evidenced-based empirical testing has not been conducted to directly assess both the construct and the assertion. Neubaum et al. (2017), using their stewardship climate scale, offer a demonstration of such testing with an example comparison of stewardship climate levels between family and nonfamily firms. Their comparison demonstrates empirically higher stewardship climate levels in family firms as compared to nonfamily firms. In effect, this study opens a wide range of research questions that heretofore have not been possible to test in both family business and the broader organizational science fields. Indeed, stewardship theory was developed to explain behavior in the broad organizational context—not solely family firms. However, family firm scholars have lead the efforts to develop a valid and reliable measure of stewardship climate that can be used across firms, both family and nonfamily. Thus, Neubaum et al. (2017) have pioneered this effort by responding to the call to strengthen the field of family business research by finding “ways to give back to these sister disciplines from which we borrow, enriching these disciplines in return” (Zahra & Sharma, 2004, p. 336).

The third article that makes a contribution to the literature by improving measurement techniques is the study by Holt, Madison, and Kellermanns (2017), which is titled “Variance in Family Members’ Assessments: The Importance of Dispersion Modeling in Family Firm Research.” As with Anglin et al. (2017), this study seeks to overcome some of the problems associated with bias when using single-source survey designs. As noted in the article, past research has often reached conclusions based on the responses of a single family member as a means to make inferences on the family firm (e.g., Eddleston, Kellermanns, & Zellweger, 2012). While an individual perspective of the firm is often appropriate, it has been acknowledged that gaining the perspective of several members of the family business, including nonfamily members, is important when asking certain research questions that reside at the family or firm level of analysis (e.g., Chrisman, Sharma, & Taggar, 2007; Kotlar & De Massis, 2013). In these circumstances, researchers often rely on shared agreement regarding a particular theoretically-derived construct and use averaged individual survey responses to capture the shared sentiments within the family firm regarding that construct.

To address the importance of differences in perceptions within family firms, Holt et al. (2017) provide an overview of the value of dispersion modeling from a theoretical standpoint and provide an example of its usage within the family firm context. Dispersion modeling (known more formally as dispersion composition models) focuses not on shared agreement but on within-group variance of a particular construct (Chan, 1998). In general, dispersion modeling is a methodological approach that seeks to capture the level of disagreement among several respondents, which makes it a useful technique for a wide variety of situations, both inside and outside of the family business domain. Specific to family business, Holt et al. (2017) provide several potential applications of this technique including, for example, how incongruent family goals or family harmony can lead to poor family-related noneconomic outcomes. Overall, the application of this technique allows for appropriate testing of research questions that have largely been unavailable to family business researchers.

Finally, the fourth article in this special issue also makes a measurement and operationalization contribution by introducing latent profile analysis (LPA), which is a technique used to reduce a large group of continuous variables into meaningful groupings. Specifically, Stanley, Kellermanns, and Zellweger (2017), in their article titled “Latent Profile Analysis: Understanding Family Firm Profiles,” demonstrate how LPA can be used to overcome some of the difficulties associated with classifying family businesses. While measuring and classifying family firms (vs. nonfamily) is a potentially important component of their article, Stanley et al. (2017) primarily contribute to the field by describing how taking a configurational approach can stimulate researchers to reconceptualize their research questions and models to be more comprehensive and interdependent. We discuss this issue in more detail in the following section.

Model Specification

Another key consideration for researchers when designing research models is appropriate specification. The development of an accurate, yet parsimonious, research model—a framework that identifies constructs and describes a particular phenomenon—is important because the model serves to test, build, or extend theory about organizations and the people who work in them. Since it is not practical or feasible to include every potential construct in a single model, researchers must often make some difficult decisions when specifying the model. In other words, model specification generally involves making theoretically driven and empirically informed trade-off decisions between inclusion and parsimony.

While it is important to avoid omitting essential variables in a model, it is also important to avoid including unneeded variables because overfitting can make the results less generalizable. Specifically for quantitative analyses, the inclusion of irrelevant variables in an analysis may even mask true effects due to multicollinearity, even if nonessential variables do not bias the model overall. For nondeductive studies, the researcher is particularly challenged by decisions regarding which constructs to give additional consideration or credence because processes may follow multiple forms of progression including converging, diverging, and parallel (Van de Ven, 2007). For both process and variance models, however, there is an expectation for explaining (i.e., theorizing) why some variables or constructs are included in the model over others, as well as how they are related (e.g., linear vs. nonlinear).

Building on this concern about parsimoniously building and testing a model, Stanley et al. (2017) demonstrate how LPA can be used to be more inclusive, while providing a more comprehensive and theoretically derived classification of firms. When compared to traditional classification approaches, including cluster analyses, the advantages of LPA become quite clear. By allowing theoretically important variables associated with a particular family business to be calculated and used, a stronger perspective can be built. Such a perspective is increasingly important in family business research because the field is steadily moving away from binomial classification schemes based on one or a few variables toward more complex, multivariate models. The more complex models are not only used to differentiate family businesses from nonfamily businesses but increasingly more often to help explain family business heterogeneity.

Conclusion

Research methodologies and analytics continue to evolve, providing increasingly sophisticated ways of analyzing organizational phenomena. As such, there is a need to continually incorporate new methods into our research. This special issue on “Process and Variance Methods in Family Business” was developed in response to this need; our desire was to provide a forum for scholars to discuss and demonstrate novel methods, measures, and techniques that can be particularly useful in the pursuit of better understanding family businesses. As the field continues to evolve, we make an explicit call for more articles of this kind, studies that explore and explain novel and useful methods and statistical techniques that can tackle the many research challenges of family business.

One particular area in need of development, which is obviously not present in our collection of articles, is that of process methods. We acknowledge that the articles contained herein largely focus on psychometric assessments of variables in variance models. None of the articles deal explicitly with how events might be better examined in reliable and valid ways. Indeed, despite our call for more process-based methods papers, the proposals we received were overwhelmingly dominated by variance-based methods. Hence, there is noticeably a need to expand our understanding of process models and methods in the future. For instance, recent concerns regarding family business heterogeneity have lead scholars to consider the different ways that family businesses differ, but events that create such differences have been largely ignored. Improvements in measuring process events are needed to understand such phenomena. For instance, developing operational decision rules for coding and measuring different events is an important area for future research in family business. Overall, our hope is that the articles contained in this special issues can serve as a platform for future work on research methods that can ultimately lead to new knowledge that affects not only family businesses but other types of organizations as well.

Footnotes

Acknowledgements

We would like to thank Keith Brigham, Mike Hitt, Nadine Kammerlander, Curt Moore, Pramodita Sharma, and Andy Van de Ven for helpful comments on earlier drafts of this article. Also, we wish to express our gratitude to the many reviewers for their developmental comments on the articles submitted to this special issue.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

Notes

Author Biographies

G. Tyge Payne is the Georgie G. Snyder Professor of Strategic Management and Jerry S. Rawls Professor of Management at the Rawls College of Business, Texas Tech University. His research interests include configurations, family business, organizational ethics, multilevel methods, social capital, strategic entrepreneurship, and venture capitalism. He currently serves as an associate editor for Family Business Review.

Allison W. Pearson is the Giles Distinguished Professor of Management and holds the Rouse Professorship in Management at Mississippi State University. Her interests include behavioral issues in family firms, including social capital, group functioning, and goals. She is currently an associate editor for Family Business Review.

Jon C. Carr is the Jenkins Distinguished Professor of Entrepreneurship at North Carolina State University. His research focuses on psychological aspects of entrepreneurship and family business, to include topics such as work–family conflict, human and social capital, and serial entrepreneurship. His additional interests include employee innovation and creativity, and the role of business model innovation within small, entrepreneurial firms. He is an associate editor for Family Business Review.

References

Anglin

A. H.

Reid

S. W.

Short

J. C.

Zachary

M. A.

Rutherford

M. W.

(2017). An archival approach to measuring family influence: An organizational identity perspective. Family Business Review, 30(1), 19-36.

Bliese

P. D.

Ployhart

R. E.

(2002). Growth modeling using random coefficient models: Model building, testing, and illustrations. Organizational Research Methods, 5, 362-387.

Bono

J. E.

McNamara

(2011). From the editors: Publishing in AMJ-Part 2: Research design. Academy of Management Journal, 54, 657-660.

Brigham

K. H.

Lumpkin

G. T.

Payne

G. T.

Zachary

M. A.

(2014). Researching long-term orientation: Construct validation and recommendations for future research. Family Business Review, 27, 72-88.

Cater

J. J.

Kidwell

R. E.

Camp

K. M.

(2016). Successor team dynamics in family firms. Family Business Review, 29, 301-326.

Chan

(1998). Functional relations among constructs in the same content domain at different levels of analysis: A typology of composition models. Journal of Applied Psychology, 83, 234-246.

Chrisman

J. J.

Sharma

Taggar

(2007). Family influences on firms: An introduction. Journal of Business Research, 60, 1005-1011.

Colli

(2012). Contextualizing performances of family firms: The perspective of business history. Family Business Review, 25, 243-257.

Daspit

J. J.

Holt

D. T.

Chrisman

J. J.

Long

R. G.

(2016). Examining family firm succession from a social exchange perspective. Family Business Review, 29, 44-64.

10.

Davis

J. H.

Schoorman

F. D.

Donaldson

(1997). Toward a stewardship theory of management. Academy of Management Review, 22, 20-47.

11.

De Massis

Sieger

Chua

J. H.

Vismara

(2016). Incumbents’ attitude toward intrafamily succession. Family Business Review, 29, 278-300.

12.

Eddleston

K. A.

Kellermanns

F. W.

Zellweger

T. M.

(2012). Exploring the entrepreneurial behavior of family firms: Does the stewardship perspective explain differences? Entrepreneurship Theory and Practice, 36, 347-367.

13.

Evert

R. E.

Martin

J. A.

McLeod

Payne

G. T.

(2016). Empirics in family business research: Progress, challenges, and the path ahead. Family Business Review, 29, 17-43.

14.

Holt

D. T.

Madison

Kellermanns

F. W.

(2017). Variance in family members’ assessments: The importance of dispersion modeling in family firm research. Family Business Review, 30(1), 61-83.

15.

Kotlar

De Massis

(2013). Goal setting in family firms: Goal diversity, social interactions, and collective commitment to family-centered goals. Entrepreneurship Theory and Practice, 37, 1263-1288.

16.

Le Breton-Miller

Miller

Steier

L. P.

(2004). Toward an integrative model of effective FOB succession. Entrepreneurship Theory and Practice, 28, 305-328.

17.

Madison

Holt

D. T.

Kellermanns

F. W.

Ranft

(2015). Viewing family firm behavior and governance through the lens of agency and stewardship theories. Family Business Review, 28, 312-331.

18.

Moorman

R. H.

Podsakoff

P. M.

(1992). A meta-analytic review and empirical test of the potential confounding effects of social desirability response sets in organizational behavior research. Journal of Occupational and Organizational Psychology, 65, 131-149.

19.

Morrison

Morgan

M. S.

(1999). Models as mediating instruments. In Morgan

M. S.

Morrison

(Eds.), Models as mediators: Perspectives on natural and social science (pp. 10-37). Cambridge, England: Cambridge University Press.

20.

Neubaum

Thomas

Dibrell

Craig

(2017). Stewardship climate scale: An assessment of reliability and validity. Family Business Review, 30(1), 37-60.

21.

Schneider

(1975). Organizational climates: An essay. Personnel Psychology, 28, 447-479.

22.

Sharma

Chrisman

J. J.

Pablo

A. L.

Chua

J. H.

(2001). Determinants of initial satisfaction with the succession process in family firms: A conceptual model. Entrepreneurship Theory and Practice, 25, 17-36.

23.

Sharma

Salvato

Reay

(2014). Temporal dimensions of family enterprise research. Family Business Review, 27, 10-19.

24.

Short

J. C.

Sharma

Lumpkin

G. T.

Pearson

A. W.

(2016). Oh, the places we’ll go! Reviewing past, present, and future possibilities in family business research. Family Business Review, 29(1), 11-16.

25.

Stanley

L. J.

Kellermanns

F. W.

Zellweger

(2017). Latent profile analysis: Understanding family firm profiles. Family Business Review, 30(1), 84-102.

26.

Van de Ven

A. H.

(2007). Engaged scholarship: A guide for organizational and social research. Oxford, England: Oxford University Press.

27.

Zahra

S. A.

Sharma

(2004). Family business research: A strategic reflection. Family Business Review, 17, 331-346.

28.

Zellweger

T. M.

Nason

R. S.

Nordqvist

Brush

C. G.

(2013). Why do family firms strive for nonfinancial goals? An organizational identity perspective. Entrepreneurship Theory and Practice, 37, 229-248.