As noted previously, Argote and Miron-Spektor (2011) developed a conception of the organizational context that includes latent and active components. The latent or background context affects learning through its effects on the active components of members, tasks, and tools. The background context determines the organization’s task and the tools available to perform its task. The background context also affects members’ abilities, motivations, and opportunities. For example, members’ abilities are affected by contextual factors such as selection methods, training programs, and performance feedback. Members’ motivations are affected by contextual factors including rewards, feedback, job design, and the organizational culture. Members’ opportunities are affected by the organization’s structure and social network.
Contextual factors that have been studied in relationship to organizational learning from direct experience are discussed in this chapter. These include the organiza- tion’s specialization, its culture, its structure, the performance feedback it provides, its training practices, resources, and power distribution. Contextual factors related to knowledge retention are discussed in Chap. 4 and those related to knowledge transfer are described in Chap. 6.
1. Specialist Versus Generalist Organizations
A dimension of the context that has been empirically examined in relationship to organizational learning is whether the organization is a specialist or generalist. Specialist organizations have been found to learn more from experience than gener- alist organizations (Haunschild & Sullivan, 2002; Ingram & Baum, 1997). For example, Barnett, Greve, and Park (1994) found that specialist banks had higher returns on assets as a function of experience than generalist banks and further that generalist banks did not evidence performance increases as a function of experi- ence. Similarly, Ingram and Baum (1997) found that “geographic generalists” that operated over a large physical area were less affected by their own experience than specialists that concentrated in a smaller number of areas. Thus, generalist organi- zations benefited less from experience than specialist organizations.
2. Organizational Culture
Another characteristic of the context that has received considerable research attention is the culture. A culture of psychological safety (Edmondson, 1999) that lacks defensive routines (Argyris & Schon, 1978) has been found to facilitate learning. When members feel psychologically safe and free to express their ideas, organiza- tions are more likely to learn from experience than when members do not feel safe. A “learning” orientation has also been shown to facilitate the relationship between experience and performance outcomes (Bunderson & Sutcliffe, 2003). When team members emphasize learning in their unit rather than comparing their unit’s perfor- mance to other units, they are more likely to learn from their experience. The shared language that members who work together develop enables the interpretation of experience (Weber & Camerer, 2003). Cohesion or liking among group members can also facilitate organizational learning (Wong, 2004).
3. Organizational Structure
The extent of decentralization in an organization has been theorized to affect orga- nizational learning. Decentralization enables an organization to explore solutions and thereby prevents it from prematurely converging on suboptimal solutions, which is especially valuable in uncertain environments (Ethiraj & Levinthal, 2004; Siggelkow & Levinthal, 2003; Siggelkow & Rivkin, 2005). Jansen, Van Den Bosch, and Volberda (2006) found in an empirical study that decentralization increased explorative innovation and had no effect on exploitative innovation.
Investigating different structures, Fang, Lee, and Schilling (2010) determined in a simulation that semi-isolated subgroups with moderate cross-group linkages promoted the greatest organizational learning. The semi-isolation of the subgroups fostered the diversity of ideas while the connections between groups fostered knowl- edge transfer across them. Bunderson and Boumgarden (2010) found that team structures characterized by specialization, formalization, and hierarchy increased team learning because they increased information sharing and reduced conflict. Jansen et al. (2006) found that formalization enhanced a unit’s exploitative innova- tion and had no effect on explorative innovation, while densely connected social relations within units enhanced both explorative and exploitative innovation. Sorenson (2003) found that interdependence engendered by vertical integration slowed the rate of learning in firms in stable environments and speeded learning in volatile environments.
4. Performance Feedback
Research on the effects of performance feedback (Greve, 2003) on organizational learning has yielded somewhat mixed results. Several researchers have documented or theorized about the positive effects of feedback. Delays in feedback have been found to hinder learning from experience (Diehl & Sterman, 1995; Gibson, 2000; Rahmandad, 2008). When members’ actions do not receive immediate rewards but occur in a sequence with an overall reward, learning can also be impaired, especially when turnover occurs (Denrell, Fang, & Levinthal, 2004). Although high-feedback specificity has been found to improve learning initially, high-feedback specificity dampens exploratory behavior over the long run (Goodman, Wood, & Hendrickx, 2004). Individual feedback has been found to amplify the negative effects of power differences on learning (Edmondson, 2002); group feedback has been found to turn the negative effects of power differences into opportunities for learning (Van Der Vegt, de Jong, Bunderson, & Molleman, 2010). In contrast to studies finding positive effects of feedback, Rick and Weber (2010) found that withholding feedback led to deeper deliberation and greater learning than providing feedback.
Training structures and processes in organizations also affect learning (Bell & Kozlowski, 2008; Ford & Kozlowski, 1996; Grossman & Salas, 2011). Two dimensions of training are especially relevant for organizational learning. One dimension is whether the training is conducted individually or in a group. Research has shown that group training is more beneficial for collective learning than individual training (Hollingshead, 1998; Liang, Moreland, & Argote, 1995). Training members of a group together promotes the development of a “transactive memory system” (Wegner, 1986), a collective system for encoding, storing, and retrieving informa- tion. Colloquially referred to as knowledge of who knows what, transactive memory systems enable the creation (Gino et al., 2010), retention (Liang et al., 1995), and transfer of knowledge (Lewis, Lange, & Gillis, 2005).
Another dimension of training systems is whether they include opportunities for members to observe experts performing tasks. Training through observing experi- enced members has been found to be more effective than training through lectures (Nadler, Thompson, & Van Boven, 2003). Through observing experts perform tasks, trainees can acquire tacit or difficult-to-articulate knowledge (Nonaka, 1991). Trainees also become members of a community and learn norms of behavior (Brown & Duguid, 1991). These advantages of observational methods contribute to the use of apprenticeship programs in a variety of professions, such as manufacturing and medicine.
6. Absorptive Capacity
Organizations that are high in “absorptive capacity” are able to recognize the value of external information, assimilate it, and apply it to develop innovations (Cohen & Levinthal, 1990). Absorptive capacity is facilitated by Research and Development activities that provide organizations with the background knowledge necessary to recognize and exploit external information (Cohen & Levinthal, 1990). Volberda, Foss, and Lyles (2010) reviewed the vast literature on absorptive capacity.
Not only do Research and Development activities facilitate learning from the experience of sources external to an organization, the activities facilitate learning from an organization’s own direct experience. Lieberman (1984) found that invest- ment in Research and Development increased the rate of learning among firms in the chemical processing industry. Similarly, Sinclair, Klepper, and Cohen (2000) found that Research and Development contributed to the productivity gains observed in a chemical firm.
7. Aspiration Levels
Aspiration levels affect organizational learning. Cyert and March (1963) theorized that when organizational performance falls below the organization’s aspiration level, search occurs and organizational change is likely. This problemistic search is typically myopic so changes resulting from it occur near the problem. Considerable empirical research has found support for the prediction that performance below the aspiration level leads to problemistic search (see Argote & Greve, 2007, for a review). Cyert and March (1963) further theorized that organizational aspiration levels adapt to the organization’s own past experience and the experience of other comparable organizations. Many empirical studies have found support for this pre- diction (Lant, 1992). Baum and Dahlin (2007) extended the behavioral theory of Cyert and March (1963) and found in their study of accidents in US railroads that as performance deviated from aspiration levels, the organizations benefited less from their own direct experience and more from the indirect experience of other firms in the industry. Desai (2008) further elaborated the behavioral theory and predicted and found that risk taking after poor performance was low when organizations had low levels of experience and poor legitimacy.
8. Slack Resources
Slack search has been theorized to affect learning and innovation as a complement to problemistic search (Cyert & March, 1963). Several empirical studies have found the predicted positive association between organizational slack and organizational learn- ing (e.g., Wiersma, 2007). Other researchers have found an inverted U-shaped rela- tionship between slack resources and innovation or exploratory search: increases in slack initially increased innovation but too much slack reduced the discipline neces- sary to produce innovations (Gulati & Nohria, 1996). Combining problemistic search and slack search, Greve (2003) found that problemistic search was more effective when organizations had a buffer of innovations generated through slack search.
9. Power and Status
Power relations within a social unit affect learning (Contu & Willmott, 2003). Bunderson and Reagans (2011) reviewed research on the effect of power and status differences on group and organizational learning and concluded that the negative effects of such differences were due to the dampening effect they had on experimen- tation, knowledge sharing, and the development of shared goals. Bunderson and Reagans (2011) further concluded that the negative effects of power and status dif- ferences could be mitigated when individuals high in the hierarchy were collectively oriented and used their power for the benefit of the group or organization.
10. Social Networks
Social networks facilitate both the search for information and its interpretation. Researchers have investigated the effects of network position, network structure, and tie strength. Ties that bridge “structural holes” or otherwise unconnected parts of a network increase exposure to information (Burt, 2004). Further, bridging ties that span structural holes are especially conducive to developing new knowledge when individuals who bridge boundaries share common third-party ties (Tortoriello & Krackhardt, 2010).
Focusing on network structures, Reagans and Zuckerman (2001) found that dense internal network structures fostered knowledge creation and transfer, espe- cially when members had specialized expertise (see also Rulke & Galaskiewicz, 2000). Focusing on network strength, Hansen (1999) found that weak ties between members facilitated the transfer of explicit knowledge, while strong ties enabled the transfer of tacit knowledge. Reagans and McEvily (2003) found that dense internal networks with links to external networks facilitated transfer over and above the effect of tie strength. Phelps, Heidl, and Wadhwa (2012) reviewed the burgeoning literature on social networks and knowledge transfer, identifying points of conver- gence and divergence.
11. Member Diversity and Stability
Two characteristics of members have been investigated in relationship to organiza- tional learning: member diversity and team stability. Several studies have examined the effect of diversity of members on organizational learning. Macher and Mowery (2003) found that team diversity moderated the relationship between experience and organizational performance in the semiconductor industry such that functionally diverse teams learned more from their experience than functionally homogeneous teams. By contrast, Ophir, Ingram, and Argote (1998) found that member diversity hindered organizational learning in Israeli Kibbutzim: diverse teams learned less from their experience than teams composed of similar members.
Several studies have examined the effect of team stability on organizational learning. Reagans, Argote, and Brooks (2005) found that team stability (the average number of times team members worked together) contributed positively to the per- formance of surgical teams. Similarly, Huckman, Staats, and Upton (2009) found that team stability was positively associated with the performance of software teams. Further, role stability or how long individuals remained in particular roles was also positively associated with the performance of software teams.
Tools can enable learning by facilitating the acquisition, storage, and sharing of information. Research on tools and organizational learning has primarily focused on information technology or knowledge management systems. Focusing on infor- mation technology, Boland, Tenkasi, and Te’eni (1994) described an information system that facilitated idea exchange in organizations. Ashworth, Mukhopadhyay, and Argote (2004) found that the introduction of an information system in a bank increased organizational learning.
Focusing on a knowledge management system, Kane and Alavi (2007) used a simulation to examine the effect of knowledge management tools, such as electronic communities of practice or knowledge repositories, on organizational learning. The researchers found that the performance of electronic communities of practice was low initially but subsequently surpassed the performance of other tools.
Empirical studies on the effect of various knowledge management systems have yielded mixed results. Based on a study of consulting teams, Haas and Hansen (2005) found a negative effect of using a knowledge management system on team performance. The more documents from a knowledge management system teams used, the worse their performance. The negative effect was stronger for experienced teams than for teams with less experience working together and stronger for teams with many rather than few competitors. By contrast, in a study of retail grocery stores, Kim (2008) found a generally positive effect of using a knowledge management system. The positive effect was particularly strong for employees in remote locations, for employees with few alternative sources of knowledge and for employees who dealt with products that did not become obsolete quickly. Thus, the repositories in knowledge management systems seem more valuable when the task is routine and employees do not have other sources of knowledge than when the task is uncertain and employees have other sources of knowledge.
New generations of knowledge management systems enabled by Web 2.0 tech- nologies have more affordances (Zammuto, Griffith, Majchrzak, Dougherty, & Faraj, 2007) than previous generations that were primarily document repositories. The knowledge that can be stored in document repositories is explicit knowledge. This knowledge can serve as pointers to who knows what, and thereby enable con- nections between members that facilitate the transfer of tacit knowledge. The con- nections, however, happen outside of the knowledge management system. Newer generations of knowledge management systems enabled by Web 2.0 technologies facilitate connections and interactions among individuals within the system through blogs and forums. These new technologies have greater affordances for transferring tacit knowledge than previous generations of knowledge management systems that primarily were document repositories. The realization of the affordances of these new technologies, however, is not automatic but rather depends on how they are used and supported in organizations.
Source: Argote Linda (2013), Organizational Learning: Creating, Retaining and Transferring Knowledge, Springer; 2nd ed. 2013 edition.