Kozlowski, S. W. J., Brown, K. G., Weissbein, D., Salas, E., & Cannon-Bowers, J. A. (in press). A multilevel approach to training effectiveness: Enhancing horizontal and vertical transfer. In K. J. Klein & S. W. J. Kozlowski (Eds.), Multilevel theory, research and methods in organizations: Foundations, extensions, and new directions. San Francisco, CA: Jossey-Bass.

ENHANCING HORIZONTAL AND VERTICAL TRANSFER 

Steve W. J. Kozlowski

Michigan State University

Kenneth G. Brown

University of Iowa

Daniel A. Weissbein

Michigan State University

Janis A. Cannon-Bowers

Naval Air Warfare Center Training Systems Division

and

Eduardo Salas

University of Central Florida

Over a quarter century ago, Terreberry (1968) predicted that organization environments were evolving to become increasingly complex, dynamic, and uncertain. As we approach the next millennium, that prediction has become reality. Organizations are increasingly pressured by technological, political, economic, social, and cultural changes that are global in scope and impact. While change is often incremental, it is also frequently rapid and unpredictable. These dynamics drive a need for continuous improvement and adaptive capabilities across all levels of the organization.

Virtually all approaches to organization improvement and development are based on enhancing the knowledge, skills, and attitudes (KSAs) or abilities of the workforce. This may be accomplished through recruitment, selection, and replacement of the workforce as one means to import upgraded abilities. Often improvement is accomplished through training to develop workforce KSAs. Organizations in the United States invest a remarkable $55.3 billion annually in training and development activities (Bassi & Van Buren, 1998). The presumption is that this sizable investment is warranted because training improves the organization’s ability to accomplish key objectives. From this perspective, training is effective to the extent that it directly contributes to the strategy, objectives, or outcomes central to organizational effectiveness (Jackson & Schuler, 1990).

Pressures for continuous improvement and adaptability place renewed emphasis on this presumed link between training outcomes and organization effectiveness. This link spans multiple levels in that training effectiveness is inherently macro, and yet it is rooted in micro level processes. Interestingly, although the contribution of training to organization effectiveness is assumed in training models (Goldstein, 1993; McGehee & Thayer, 1961), the specific nature of this linkage is not well articulated. Moreover, this critical linkage has received relatively little research attention. The best current efforts to establish training effectiveness end with demonstrating transfer of individual-level training outcomes to the job context. There is a presumption that individual-level training outcomes aggregate and emerge to create valued outcomes at higher levels, but there is little theoretical guidance as to how to conceptualize and model these effects. Hence, there have been few efforts to do so.

This situation creates a levels paradox. The primary goal of training is to enhance organizational effectiveness. That is, training is predicated on contributing to higher level group and organization objectives, outcomes, and results. Yet, the models, methods, and tools of training are focused on the individual level. Training needs are derived for individual level KSAs, training programs are delivered to individuals, and training effectiveness is evaluated for individual-level outcomes and transfer. The organizational context, its effects on training, and the means by which training effectiveness will impact organizational objectives are not adequately addressed in current models. Thus, there is a levels gap between theoretical models of training needs assessment, design, and evaluation, and the higher levels at which training must have an impact if it is to contribute to organizational effectiveness (Kozlowski & Salas, 1997).

The purpose of this chapter is to explicate a multi-level model that bridges this gap. We focus on training transfer because it is the primary leverage point by which training can influence organizational effectiveness. Levels issues are embedded in two distinct transfer foci: horizontal and vertical transfer. Horizontal transfer (Kozlowski & Salas, 1997) refers to transfer across different settings or contexts at the same level. Horizontal transfer has been the primary focus in traditional models of training effectiveness. Vertical transfer (Kozlowski & Salas, 1997) refers to transfer upwards across different levels of the organizational system. Vertical transfer is concerned with the link between individual training outcomes and outcomes or results at higher levels of the organizational system.

Qualitatively distinct levels issues are relevant to each of these transfer foci. One issue concerns top-down contextual effects (Kozlowski & Klein, this volume) that facilitate or inhibit individual learning and transfer. Top-down contextual effects are group and organizational factors that have direct and moderating effects on learning and transfer. These effects have been the source of recent theory and research addressing the influence of organizational factors on motivation to learn, transfer, and training effectiveness at the individual level of analysis. Top-down effects are relevant to horizontal transfer. The second issue concerns bottom-up emergent processes (Kozlowski & Klein, this volume) that influence the contribution of individual level training outcomes to higher level group and organization outcomes. This second issue is concerned with strengthening the linkage between training and organization effectiveness. The processes by which individual level training outcomes combine and emerge, bottom-up, to manifest as higher level outcomes has been largely neglected in training theory and research. We acknowledge that experienced practitioners may consider vertical transfer processes, at least implicitly. However, the link between individual and higher level training outcomes has been neglected in formal research and practice models. Bottom-up effects are relevant to vertical transfer. In our view, vertical transfer is a key leverage point for strengthening the link between training and organizational effectiveness.

An example may be useful for illustrating the importance of both transfer foci. Consider Total Quality Management (TQM) programs that are intended to contribute to organizational objectives of continually improving processes, products, and services. In many instances, TQM training is the primary intervention for accomplishing these organizational objectives. Yet, even if individuals learn the trained knowledge and skills, there are often many incongruent contextual factors in the job setting that are not addressed as part of the development effort. Top-down contextual factors, such as leadership, structure, job design, rewards, climate, and so forth, can inhibit the horizontal transfer of outcomes when they are not aligned with the content of training. When there is no support from one’s leader or peers, insufficient opportunity to practice new skills, no rewards, and a climate that is inconsistent with the trained skills, horizontal transfer of the trained knowledge and skills to the work setting is unlikely. In the absence of horizontal transfer, training cannot contribute to organizational effectiveness.

Furthermore, the typical TQM program is designed to deliver the exact same training content to virtually every member of the organization. Everyone is responsible for quality, and everyone is exposed to the same training content. With respect to vertical transfer as an emergent process, the implicit assumption of this approach is that improving the "quality" of organization products, processes, and services is an additive function of similar individual outcomes or contributions: Each individual contribution sums together to yield higher level impacts. Training will contribute to organization effectiveness when this assumption holds. That is, when quality is really a sum of incremental contributions, this approach will have impacts on quality outcomes at higher levels in the system. Yet, in many (if not most) organizations, improving quality for higher level objectives is not consistent with a simple additive model. Rather, quality is a complex combination of different contributions across individuals and across different levels of the organization. When this more complex model holds, the typical approach to TQM training will be, at best, inefficient. More likely it will be ineffective. The distinctive knowledge and skills needed for different individuals and different levels, and the integration of these skills, will be neglected by training. If training is to meet its promise of improving the organization’s capability to learn, adapt, and improve, these level gaps must be bridged.

We address both levels issues in this chapter, which is structured as follows. First, we consider top-down contextual influences on horizontal transfer. We assert that the foundation of theory, methods, and research findings is sufficiently well-developed such that top-down contextual effects should be explicitly incorporated in training research and practice. We review this foundation and use it to specify basic levels principles. That is, we regard the principles as established knowledge. These principles are intended to promote state-of-the-art research and practice that enhances the horizontal transfer as one aspect of training effectiveness. Second, we assert that neglected bottom-up processes should be the focal point for theory development and new research to elaborate the linkage between training and emergent outcomes at higher levels in the organization system. This is intended to bridge the gap between training outcomes and organization effectiveness. We present a theoretical framework to guide new research on vertical transfer. The framework distinguishes two distinctive forms of emergence -- composition and compilation -- that characterize different types of vertical transfer processes. We describe distinguishing characteristics and present examples to illustrate different types of emergent effects. The model is then used to posit theory-based propositions regarding vertical transfer that have implications for training needs assessment, design / delivery, and evaluation. In other words, propositions represent theoretical assertions in need of research evaluation. Our intent is to promote conceptual development, influence training research, and suggest potential implications for practice. By doing so, we hope to stimulate research on vertical transfer, a neglected and yet critical aspect of training effectiveness.

END OF INTRODUCTION