A Meta-Analytic Integration of What Matters in Training Transfer

Estimates demonstrate that 52- 92% of acquired learning is lost within a year following training (Arthur, Bennett, Stanush, & McNelly, 1997; Saks, 2002), wasting billions in organizational spending on training each year (Miller, 2012, 2013, 2014). As such, research on training transfer has garnered attention from theoretical and empirical research alike (e.g., Baldwin & Ford, 1988; Blume, Ford, Baldwin, & Huang, 2010; Ford & Weissbein, 1997; Tracey, Tannenbaum, & Kavanagh, 1995) to better understand the factors which enhance the process of training transfer. Among the various factors that have been identified as important, factors of the work environment have received much attention in the recent research. In fact, empirical work has shed light to the roles of organizational support and motivation to transfer in predicting training transfer. Beyond this basic understanding, research is needed to explore the nature of transfer in different evaluation contexts and the differential effects of various levels of support. Thus, the current dissertation uses meta-analytic techniques to examine the extent to which four factors of work environment support predict training transfer as it differs in context. First, motivation to transfer, organizational support, supervisor support, peer support and opportunities to perform all correlate moderately and positively with training transfer (ρ=0.15-0.38); interestingly, the nature of the relationships between work environment characteristics, motivation to transfer, and training transfer does not appear to differ significantly even when transfer is evaluated a year following training (ρ=0.25-0.57), yet are based on low k. Second, motivation to transfer was found to fully mediate two relationships- organizational support and peer support- to training transfer. Interestingly, although not explained by motivation to transfer, supervisor support explains the most variance (i.e., 31% of R) of work environment support factors in explaining transfer. Moderator analyses attempted to explore the impact of transfer task, industry type, and timing of the predictor assessment in relation to training; however, insufficient k was reported for fair comparisons to be made across groups. Ultimately, this study aims to inform theory and impact the state of the science such that practitioners can feel confident that the time and effort spent in ensuring training transfer is well-spent