Proposal view
Proposal Type: Individual Thematic Poster 
Domain: Lifelong Learning and Professional Development 
SIG: Learning and Professional Development 
Equipment  
Paper Details
Title The influence of individual-, training- and environment - related factors on training transfer
Abstract
The transfer of newly learned knowledge and skills into new situations has been largely studied. In the area of professional development, it has been showed that little of what was learned during training programs was subsequently transferred to the workplace. A critical issue is therefore to investigate the factors that are likely to foster or hinder transfer. The Learning Transfer System Inventory (LTSI; Holton, Bates, & Ruona, 2000) considers 16 factors likely to influence the transfer of training to the workplace. The purpose of this study was to translate the LTSI into French and to examine (1) the internal structure of the translated instrument, and (2) its predictive validity regarding transfer 1 to 3 months after training. First, a factor analysis revealed a factor structure very similar to the original structure: the 15 of the original factors were replicated. Second, 8 factors displayed statistically significant correlations with transfer: Information given to the trainees before the training, Similarity between the training and the job, Centrality of the transfer issues during the training, Expectancy that transfer will lead to a better performance, Expectancy that this increased performance will be rewarded, Motivation to transfer, Self-efficacy toward transfer, and Resources availability. The results also show that the factor Motivation to transfer and Self-efficacy mediate the relation between these variables and transfer. Finally, they are discussed and integrated into Eccles and Wigfield’s expectancy-value model.

Summary

The transfer of newly learned knowledge and skills into new situations has been the focus of many discussions and debates (e.g. Ellis, 1965; Perrenoud, 1997; Tardif, 1999). In the lifelong learning and professional development area, it has been showed that little of what was learned during training programs was subsequently transferred to the workplace (e.g. Alliger & Janak, 1997). For example, teachers who attend a training program on peer-assisted learning are likely to use only a small portion of what they learn in their classroom. What are therefore the factors likely to increase training transfer? Empirical support was found for variables such as peer and supervisor support (e.g. Facteau, Dobbins, Russell, Ladd, & Kudisch, 1995), organizational culture (Tracey, Tannebaum & Kavanagh, 1995), training reputation (Facteau et al., 1995), goal-setting posttraining intervention (Gist, Stevens, & Bavetta, 1991), self-efficacy, locus of control, and conscientiousness (Colquitt, LePine, & Noe, 2000).

Further, instruments which assess a whole set of relevant and non-redundant factors have also been developed (e.g. Holton, Bates, & Ruona, 2000; Tracey & Tews, 2005). Holton et al. (2000) developed the Learning Transfer System Inventory (LTSI). This instrument measures 16 factors likely to influence transfer. Three of those factors evaluate to the quality of the training program (Amount of information given to the trainees before the training, Similarity between training and job situations, and Centrality of the transfer issue during the training); five are related to social support (Colleagues and Supervisor support vs. opposition to transfer, Performance coaching, and Openness to change of the workteam); four focus on rewards (Rewards resulting from transfer, Reprimand resulting from non-transfer, Expectancy that transfer will lead to a better performance, Expectancy that this increased performance will be rewarded); two factors focus on resources availability (Time and Material/financial/human resources); and two tap into motivational variables (Motivation to transfer and Self-efficacy toward transfer).

Several aspects of the validity of the instrument (internal structure, predictive validity and cross-cultural validity) have been successively tested (e.g. Bates, Holton, Seyler & Carvalho, 2000; Chen, Holton, & Bates, 2005; Khasawneh, Bates, & Holton, 2006; Yamnill, 2001).

The purpose of this study is therefore, on the one hand, to translate the LTSI into French and to test its internal structure, and, on the other hand, to use the instrument to predict effective transfer to the workplace.

 

Method

 

The LTSI (68 items) was completed by 328 participants who attended a training program (58.6% of males; mean age = 39.4). One to 3,5 months later, these participants were administered the transfer questionnaire (10 items), which is a self-reported measure of their level of transfer. 106 of the 328 participants completed this second questionnaire. Both questionnaires were sent by e-mail, anonymous, and the matching between the two was done with an anonymous code.

The factor structure of the LTSI was investigated with a factor analysis (Maximum likelihood) and an oblique rotation (Oblimin). Selection criteria for factors and items were an eigenvalue greater than 1 and a loading greater than .40, respectively. The predictive validity was investigated with correlations, and the mediation analyses were computed with regression analyses (see Baron & Kenny, 1986) and Sobel tests.

 

Results

 

On the one hand, in regard to the internal structure of the LTSI, 15 of the 16 original factors emerged; only the factor Performance coaching was not replicated. The model explains 66.8% of the variance, Cronbach alphas range from .67 to .92 (mean = .79). Four items were eliminated because they displayed insufficient loadings, and three items moved from one scale to another.

On the other hand, in regard to the correlation with transfer, eight factors displayed statistically significant correlations with transfer:

-                     the three training-related factors : Information before training (r = .21, p < .05), Similarity between the training and the job (r = .30, p < .01), and Centrality of transfer (r = .35, p < .001)

-                     two factors related to Vroom’s (1964) Value-instrumentality-expectancy model : Expectation of performance improvement after training (r = .26, p < .01), and Expectation of rewarding of the performance (r = .27, p < .01)

-                     the two individual variables : Self-efficacy (r = .25, p < .01), and Motivation to transfer (r = .46, p < .001)

-                     and Resources availability (r = .25, p < .01).

Further, mediation analyses showed that the factor Motivation to transfer was a complete mediator of all the factors significantly related to transfer; and Self-efficacy was a partial mediator of the factors Resources availability and Expectation of performance improvement after training, and a complete mediator of the factors Information before training and Expectation that the improved performance will be rewarded/valued.

 

Discussion

 

This study shows encouraging results toward the understanding of the factors likely to influence the transfer of training to the workplace. If we go back to the example of the training program on peer-assisted learning, this study suggests that the teachers that attended this training would be more likely to transfer their knowledge if: they received information about the training program before it began (e.g. content, objectives), the issue of transfer was largely dealt with during training (e.g. example of when and how to use the training, discussions on the possible obstacles to transfer and ways to overcome them), they expect that transfer will help them to do their work better and that this improvement will be noticed and valued, and they feel able of using their new competences in their job, even in the face of obstacles.

Further, the content of the two mediating variables (Motivation to transfer and Self-efficacy) refer to two of the Eccles and Wigfield’s (cfr. Wigfield & Eccles, 2002) expectancy-value model: utility and expectancy. The results of this study can therefore be framed in their model like is showed in Figure 1, and suggest future research directions in investigating the antecedents of Eccles and Wigfield’s (2000) constructs of utility and expectancy.
Keywords Adult education/development
Workplace learning
Appendices Figure 1 - A model of the antecedents of transfer.doc 
Authors
Name Surname Institution Country e-mail EARLI Number Presenting
Christelle Devos Catholic University of Louvain Belgium christelle.devos@psp.ucl.ac.be   *  
Xavier Dumay Catholic University of Louvain Belgium xavier.dumay@psp.ucl.ac.be    
Visit NQcontent
© European Association for Research on Learning and Instruction, 2012 All rights reserved.