Motivation to transfer revisited

Should the construct ‘motivation to transfer’ used in human resource development and management research be also used in learning research? The current study revisited motivation to transfer on a sample of 128 participants of occupational health training. Confirmatory factor analysis and partial least squares-based path modeling were used to test the hypothesized dimensions and relationships among variables including social and affective cues on training transfer. Based on a combination of the theory of planned behavior, expectancy theory, and self-determination theory, we validated three dimensions of transfer motivation: autonomous motivation to transfer, controlled motivation to transfer, and intentions to transfer. Results indicate that autonomous motivation was affected by attitudes toward training content and utility reactions; controlled motivation was affected by utility reactions, supervisory support, and social norms. Intentions to transfer mediated the effects of autonomous motivation on transfer three months after training. Implications of a multidisciplinary perspective combining learning sciences and human resource development are discussed. The Noe Model Revisited In 2006, in a discussion on how motivation influences transfer, Pugh and Bergin (p. 155) recommended adopting the construct motivation to transfer from the human resource development (HRD) literature to the learning sciences; their recommendation was published in a highly influential journal in the field, Educational Psychologist. Twenty years earlier, Raymond Noe coined the term ‘motivation to transfer’ that Pugh and Bergin were referring to as “the trainees’ desire to use the knowledge and skills mastered in the training program on the job” (Noe, 1986, p. 743). His work was published in a highly influential journal in the field of management and HRD, Academy of Management Review. It is argued here that it would be fatal if learning scientists followed Pugh and Bergin’s advice to adopt the construct from HRD research to learning research. This is for five reasons. First, motivation to transfer currently is a one-dimensional construct that is too coarse to adequately reflect motivational states and traits in the transfer process. Second, motivation to transfer currently is not, and was not, adequately grounded in motivation theories. Third, there is poor empirical evidence in the HRD and management literature on how motivation to transfer affects transfer of learning. Fourth, motivation to transfer currently is a static construct, only measured immediately following an intervention. Finally, Noe (1986) said little in his original conceptualization about the social origin of transfer motivation; this reflected the zeitgeist of the 1980s epoch. Learning researchers now can be thoughtful in adopting a more balanced view. It is argued further that motivation to transfer does indeed play a central role in the transfer process; there is enough theoretical (and common sense) backdrop to expect that learners who are motivated to apply training content from training to work, or from the classroom to out-of-classroom use, are more likely to be successful than learners who are not motivated. Yet, in its current form, motivation to transfer seems to be a construct powerless to account for intentional transfer processes. Transfer, from a human resource perspective, is defined as the successful application of newly trained knowledge, skills, and attitudes to the workplace (Noe, 1986; Burke & Hutchins, 2007). Hence, we argue that before following Pugh’s and Bergin’s recommendation, Noe’s model of motivation to transfer must be revisited in several significant respects: dimensionality, theoretical grounding, empirical evidence, measurement time, and social nature of motivation. On revisiting dimensionality, motivation to transfer has been measured as a one-dimensional construct in empirical training research over the past twenty-five years or so (Noe & Schmitt, 1986; Velada, Caetano, Bates, & Holton, 2009). In our recent review on motivation to transfer (Gegenfurtner, Veermans, Festner, & Gruber, 2009), we found that all 31 studies on transfer motivation reported in international peer-reviewed journals from 1986 to now used a one-dimensional scale to assess the construct, ranging from one item to eleven items. This is a paradox since motivation researchers have explored numerous dimensions of motivational processes in human actions. To name just a few, we know of intrinsic and extrinsic motivation; conscious goal intentions and unconscious implementation intentions; expectancies, instrumentalities, and valences; various types of motivational regulation and mindsets; and we differentiate motivation, volition, and emotion. All contribute to our understanding of the many facets and colors that motivation has. Motivation to transfer ICLS 2010 • Volume 1 453 • © ISLS potentially goes in concert with all of the dimensions just mentioned. It can be argued that motivation to transfer has remained a one-dimensional construct to date because it lacks a clear grounding in motivation theories. On revisiting theoretical grounding, Noe (1986; Noe & Schmitt, 1986) made no explicit statement as to which theoretical framework the idea of motivation to transfer is based; although there are references to Bandura, Latham, and Vroom in their original papers, the associated motivation theories refer not to transfer motivation but to other aspects of the papers. Hence, motivation to transfer lacked from the start, and still lacks, a solid foundation in motivation theories. In an attempt to provide a first step toward resolving this situation, we conducted two studies in which the theory of planned behavior (Gegenfurtner & Gruber, 2009) and expectancy and self-determination theory (Gegenfurtner, Festner, Gallenberger, Lehtinen, & Gruber, 2009) were tested as a basis for transfer motivation scales. In short, these studies indicate that motivation to transfer has several dimensions: autonomous motivation to transfer, controlled motivation to transfer, and intentions to transfer. These are the first steps, and more efforts are needed to validate the construct’s theoretical grounding for human resource development beyond. On revisiting empirical evidence, it is a likely explanation that the one-dimensionality and the lack of theoretical grounding account for the current state of poor evidence on the intention-behavior relationship in training transfer. Correlation coefficients ranging from 0.04 (Tziner, Haccoun, & Kadish, 1991) to 0.63 (Machin & Fogarty, 1997) suggest that the relation between motivation to transfer and transfer of training needs further elaboration. Again, we believe that a starting point for this elaboration is a reconceptualization of Noe’s model of transfer motivation. This may help to form new research practices with respect to the multidimensional nature of transfer motivation, its groundwork in valid motivation theories, and its measurement time. On revisiting measurement time, our review of transfer motivation (Gegenfurtner et al., 2009b) indicated that, in 30 of 31 cases, the construct was assessed at the immediate end of training. Only Leitl and Zempel-Dohmen (2006) measured transfer motivation at the immediate end and at a later point in time, three months after the intervention. We believe that motivation to transfer in particular, like motivation for action in general, dynamically changes over time. Hence, assessment of transfer motivation at a time when the learner has had time to explore opportunities to use training in out-of-classroom applications may provide a very different picture than the assessment of transfer motivation at a time when the learner is just about to leave the classroom. Yet, empirical examinations at a different time than at the immediate end of training are almost non-existent. On revisiting the social nature of motivation, there is large consensus that motivation is intrinsically social in nature (Hickey, Moore, & Pellegrino, 2001; Järvelä, Volet, & Järvenoja, 2010; Vauras, Salonen, & Kinnunen, 2008). We argue that this holds equally for transfer motivation. Especially when trainees spent some time at the workplace and could test the training content about its usefulness, social cues might affect formation and persistence of motivational states to transfer training. Arguably, the generation of more stable motivational traits is also influenced by factors such as social norms, feelings of relatedness, or control beliefs that are based on environmental (working) conditions emanated in social interaction. Noe (1986) constructed his model of transfer motivation at a time in which the social nature of motivation was not on the agenda yet. Reconceptualizing the Noe model can account for this development in motivation research. Present Study The present study had a dual goal. First, it aimed to test the multidimensionality of motivation to transfer. Multidimensionality was achieved by using a combined motivation theory approach. Specifically, there was a combination of a validated framework from educational psychology, self-determination theory (Baard, Deci, & Ryan, 2004), and a validated framework from management research, expectancy theory (Vroom, 2005), to conceptualize autonomous and controlled motivation to transfer. Expectancy theory as a cognitive-choice approach and self-determination theory as a need-motive-value approach can complement each other in predicting and explaining human performance in the workplace (Kanfer, 1990). Thus, instrumentality and valence items reflecting externally prompted reasons to transfer were used to assess controlled motivation. Conversely, instrumentality and valence items reflecting internally regulated behaviour were used to assess autonomous motivation. Autonomous motivation to transfer is defined here as an internalized desire to transfer learning that is initiated and governed by the self (i.e., regulated by identification or by integration with one’s values), and controlled motivation as a desire to transfer learning that is not initiated and governed by the self (i.e., regulated by external rewards or sanctions). In addition to autonomous and controlled motivation as measures of motivational traits, intentions to transfer are also included as measures of motivational states. This is because contrary to motivational traits, intentions represent a more activated, situation-specific motivational state. We used Ajzen’s (1991) theory of planned behaviour, a well-validated framework from social psychology, to conceptualize intentions to transfer. The second goal was to test a hypothesized path model, shown in Figure 1. The model explored a motivational sequence. Effects of autonomous and controlled motivation on training transfer are mediated by the more situation-specific state of transfer intentions; at the same time, situational but distal social (relatedness, support, control, and norms) and affective (attitudes and utility reactions) cues on transfer intentions are ICLS 2010 • Volume 1 454 • © ISLS mediated by more stable motivational traits, i.e., autonomous and controlled motivation to transfer. It can be argued that motivational traits act like filters. For example, supervisory support may or may not lead to activated intentions, depending on the motivational trait. The model has a range of hypothesized relationships; these are based on theory or past empirical evidence. Specifically, the hypothesis, based on self-determination theory, predicted that relatedness (RE) at work would foster internalization of external regulations and would thus be positively related to autonomous motivation to transfer (Baard et al., 2004; Gegenfurtner et al., 2009a). Supervisory support (SU) was shown to have mixed empirical results on motivation, depending on whether support was perceived to be instrumental or not; we thus expected relations to both autonomous and controlled motivation. Based on the theory of planned behavior (Ajzen, 1991), it was hypothesized that perceived behavioral control (PBC) promoted trainees’ feelings of autonomy and would thus be more important for autonomous motives; at the same time, social norms (NO) were hypothesized to have stronger effects for those trainees with controlled motivation. Past research showed that attitudes toward training content (AT) were related to both autonomous and controlled motivation (Gegenfurtner et al., 2009a). Utility reactions (UT) were hypothesized to affect self-determined feelings of autonomy; however, due to their instrumental nature, it was speculated that they could also affect controlled motivation. Figure 1. Hypothesized path model.