Multi-Level Issues in Organizational Behavior and Strategy: Volume 2

Fred Dansereau, Ph.D., is a Professor of Organization and Human Resources in the School of Management at the State University of New York at Buffalo. He received his Ph.D. from the Labor and Industrial Relations Institute at the University of Illinois with a specialization in organizational behavior. Dr. Dansereau has extensive research experience in the areas of leadership and managing at the individual, dyad, group, and collective levels of analysis. Along with others, he has developed a theoretical and empirical approach to theorizing about and testing theories at multiple levels of analysis. He has served on the editorial review boards of the Academy of Management Review, Group and Organization Management, and Leadership Quarterly. Dr. Dansereau is a Fellow of the American Psychological Association and the American Psychological Society. He has authored six books and over 70 articles and is a consultant to numerous organizations, including the Bank of Chicago, Occidental, St. Joe Corp., Sears, TRW, the United States Army and Navy, Worthington Industries, and various educational institutions.Francis J. Yammarino, Ph.D., is a Distinguished Professor of Management and Director and Fellow of the Center for Leadership Studies at the State University of New York at Binghamton. He received his Ph.D. in Organizational Behavior (Management) from the State University of New York at Buffalo. Dr. Yammarino has extensive research experience in the areas of superior-subordinate relationships, leadership, self-other agreement processes, and multiple levels of analysis issues. He serves on the editorial review boards of seven scholarly journals, including the Academy of Management Journal, Journal of Applied Psychology, and the Leadership Quarterly. Dr. Yammarino is a Fellow of the American Psychological Society and the Society for Industrial and Organizational Psychology. He is the author of seven books and has published over 80 articles. Dr. Yammarino has served as a consultant to numerous organizations, including IBM, Textron, TRW, Lockheed Martin, Medtronic, United Way, and the U.S. Army, Navy, Air Force, and Department of Education.

“Multi-Level Issues in Organizational Behavior and Strategy” is Volume 2 of Research in Multi-Level Issues, an annual series that provides an outlet for discussion of multi-level problems and solutions across a variety of fields of study. Using a scientific debate format of a key scholarly essay followed by two commentaries and a rebuttal, we present in this series theoretical work, significant empirical studies, methodological developments, analytical techniques, and philosophical treatments to advance the field of multi-level studies, regardless of disciplinary perspective.

Recent years have seen an upsurge of interest in the study of emotions in organizations. Research, however, has been hampered by the ephemeral nature of emotions and a lack of an integrated multi-level model. This article therefore presents a five-level model of emotions in organizations. At the lowest level is within-person variation, defined in terms of affective events theory. Levels of the model then proceed through individual, dyadic relationship, group, and organization-wide perspectives. The article also outlines the neurophysiological processes that underlie the experience, perception, and communication of emotion; it concludes with a discussion of implications for research and practice.

This paper argues for the integration of affect, cognition, and motivation, and suggests that this integration can be especially helpful in understanding organizational behavior, a point developed also, but in a slightly different way, by Ashkanasy in his recent paper, “Emotions in organizations: A multi-level perspective.” In addition, this paper points out that recent research has made clear that it is a misconception to think that positive affect typically leads to heuristic processing and interferes with systematic cognitive processing.

Ashkanasy has presented a model and review of multi-level analyses of emotions in organizations. In this paper, the strengths and shortcomings of Ashkanasy’s paper are discussed. Chief among the strengths are the analyses of core biological processes, the discussion of within versus between person analyses, and the distinction drawn between moods and discrete emotions. A limitation is the relative lack of process integrations among levels. The nature of such integrations and the difficulty in developing them are discussed briefly in this paper.

Weiss and Isen have provided many supportive comments about the multi-level perspective, but also found limitations. Isen noted the importance of integrating affect, cognition, and motivation. Weiss commented similarly that the model lacked an integrating “thread.” He suggested that, to be truly multilevel, each level should constrain processes at other levels, and also provide guidance for the development of new concepts. Weiss also noted that the focus on biological processes was a strength of the model. I respond by suggesting that these very biological processes may constitute the “missing” thread. To illustrate this, I discuss some of the recent research on emotions in organizational settings, and argue that biology both constrains and guides theory at each level of the model. Based on this proposition, I revisit each of the five levels in the model, to demonstrate how this integration can be accomplished in this fashion. Finally, I address two additional points: aggregation bias, and the possibility of extending the model to include higher levels of industry and region.

We argue in this chapter that quantitative techniques alone cannot enable appropriate verification of the levels in which a construct should be defined and validated. We propose a process model that takes advantage of qualitative and quantitative methods when specifying and measuring the level(s) of analysis in construct development. The model implies that operationalizing a construct and its measurement should involve both qualitative and quantitative methods. Level specification begins with induction of constructs and continues with qualitative or quantitative operationalizations that are verified through the convergence of using triangulation and aggregation tests. Construct development continues until the level of analysis is accurately specified. We provide a theoretical example of construct development from the leadership literature and an empirical example to represent the verification process.

This comment provides a discovery-oriented reading of grounded theory which supplements the verification-oriented approach to level specification. I address how grounded theory can be used to discover new constructs, to surface properties of constructs, to validate constructs, and to enhance understanding of levels of analysis of constructs. Two approaches to the integration of qualitative and quantitative data in grounded theory are discussed: the linking approach and computer-aided interpretive textual analysis. The comment shows how multi-level constructs can be developed from real life interaction using discovery oriented grounded theory.

The grounded theory method is more appropriate than just a grounded theory “approach,” for teasing out the detail of the level of analysis of constructs. Also, triangulation is important to this kind of research, but the methodological distinctions between qualitative and quantitative data and qualitative and quantitative analysis need to be made clear when mapping out a methodology. The contention here is that the qualitative analysis of quantitative data is more important than the quantitative analysis of qualitative data. Qualitative analysis in line with the full grounded theory method will generate explanations of how and why a construct is represented at various levels of analysis. In turn, such an explanation can illustrate whether the questionnaire instrument is representing the levels of analysis of the construct adequately or not.

The quantitative literature on levels of analysis emphasizes identifying the level of a construct as a confirmation rather than a discovery process. Based on the comments we received from Parry and Gephart we integrate some of their ideas regarding discovery in our earlier model to augment the verification process associated with construct validation. However, we also emphasize the need to use structure even with qualitative methods, given the nature of the level verification process. We begin by discussing their specific comments as well as our reactions to these comments. We then discuss the ideas of theoretical sampling, discovery and levels of abstraction, and include an example to demonstrate the utility of using a “discovery”-oriented approach to level specification for construct validation. Finally, we provide a brief conclusion for our reply.

The resource-based view of the firm and strategic groups research are two of the most investigated frameworks in strategic management. Historically, assumptions behind these two views have seemingly put them at odds. The resource-based view of the firm argues that sustained competitive advantage is best attained when firms have unique resources, while strategic groups research argues that a number of firms within the same industry can achieve sustained profitability with strategies that are similar to one another, but distinct from other industry members. The two views focus on different levels of analysis and each largely ignores the other’s focal level. Yet neither offers any propositions that are incompatible with the tenets of the other. Thus, conceptual integration that crosses levels of analysis is possible and potentially fruitful. Indeed, some strategic groups research has begun to bridge the gap between these two theories by suggesting that firm differences exist both within and between strategic groups. This article adopts a multi-level view by developing propositions concerning contingencies when firm differences, group processes, or both may lead to sustained competitive advantage. Implications for practitioners as well as suggestions for future theory building and empirical tests are also discussed.

Integration of the Resource-Based View of firm performance and the findings from Strategic Groups Research must allow for the separation of firm level effects from those of membership in a particular strategic group. These perspectives have focused on differing levels of analysis. Further integration requires combining variables from these different levels of analysis, recognizing the MESO nature of both the relationships among these variables as well as the fundamentally inter-level character of these variables themselves. It is shown that the meaning and measurement of these terms, as well as the statistical procedures used to test relationships among them, must be better understood in order to extend an initial integration of the RBV and SGR provided by Short, Palmer and Ketchen (2003). A framework for such an extension is discussed.

Strategic group theory has two distinct aspects. The first is a resemblance among firms grounded in a common pattern of investment; the second is a mutual dependence among firms in the strategic group. Each aspect is analyzed to develop implications for research generally and multi-level research particularly.

The commentaries offered by Professors Joyce and Michael on our chapter, “Multilevel influences on firm performance: Implications from the resource-based view and strategic groups research,” provide a number of suggestions for incorporating levels of analysis into studies of firm performance. In this rejoinder to their comments, we note the need to include managers’ cognitions in studies, emphasize the role of theory in studying strategic groups, describe the value of creating conceptual clarity regarding strategic groups, and offer suggestions for studies that might emanate from the ideas in the commentaries in juxtaposition to our original work.

We introduce a multi-level model of the dependence of interfirm network topologies on the distribution and commonality of information in a network and the information strategies pursued by its member firms. Network topology, information properties of the network, and firm-level action within the network form dynamic, recursive, cross-level relationships – information properties in the network determine firm-level action, which in turn impacts the network topology and information properties. We derive predictions about the kinds of information strategies that firms are likely to adopt and succeed with in different information regimes, and about the kinds and short- and long-run dynamics of network topologies expected under different information regimes. Our model sheds new light on network topologies as a dependent variable that can be explained by network-level information regimes and firm-level information strategies.

The Moldoveanu, Baum & Rowley approach highlights the importance of developing a multi-level theory of network evolution that explicitly accounts for strategic agency and (bounded) rationality. I discuss intriguing questions raised by their model about structure vs. topology, feedback loops in evolution, and network resources other than information.

This commentary addresses the problem of interfirm network formation from the perspective of multiple types of relationship content and of network structure. The approach builds on Burt’s (1980) typology of network structures and on a range of empirical studies on interorganizational networks. The chapter by Moldoveneau, Baum and Rowley on network evolution from an information-sharing perspective captures part of this research domain. The challenges posed by network evolution research are discussed in the broader light of multilevel analysis.

In this reply, we respond to a series of substantive comments on our work by both Madhavan and Walker. In our response to Madhavan’s comments, we consider three accounts – “weak,” “semi-strong” and “strong” – that clarify how our model “explains” and offers insights that can emerge from our modeling strategy. We also explore ways in which our explanatory strategy might be extended in light of his critique. In our response to Walker’s comments, we adopt the “semi-strong” thesis, which admits variation in network-generating mechanisms, while also recognizing that information needs to be distributed and shared in order for many types of networks to function.

Drawing upon prior work in cognition, organizational learning, strategy, and organization theory, a multi-level, longitudinal account of change in core competence is presented. The central concern addressed is how knowledge passes from “individual” to “organizational,” and how insight present in one part of the organization comes into broader currency. Data from a mature industry firm form the foundation for a depiction of change as a cognitive process involving multiple individuals, groups, and levels within the organization. Speculative propositions and a process model are presented.

The commentary discusses connections between Jelinek’s analysis of strategic change and research on organizational learning. Points of convergence and divergence between the strategic change case and the organizational learning literature are identified in three areas: the role of experience in organizational learning, knowledge transfer within and between organizations, and the embedding of individual knowledge in organizational memory systems and supra-individual routines. Directions for future research are suggested.

Strategic change and managerial cognition are closely intertwined. However when data on a specific case of delayed strategic change involving multiple levels of an organization are presented, questions are raised about the application of a managerial cognition framework. This commentary attempts to reconcile some of what is currently known regarding managerial cognition and how a cognitive model is induced from a case study on strategic change in the James Division of Warren Corporation. Alternative interpretations of the data are introduced and directions for future research using the induced theory are suggested.

A cognitive view of implementing strategic innovation sees individual learning of senior managers became organizational only when Operating Logic shifted to Innovating Logic at the operating level, which required recognition at the operating level that change was needed. Organization learning concepts extend the account, while simpler explanations of entry threat and changing division management collapse organizational cognition to top-management, directing attention away from proactive strategic innovation. The cognitive view emphasizes broader cognitive resources and participation to enable organizational learning by doing.