1. Introduction
The importance of teachers’ in-service training for COVID-19 prevention guidance at the school setting is well established for students and community safety [
1,
2]. Teacher in-service training for epidemic control equips teachers with the resources, skills, and assistance they need to properly carry out their profession’s tasks, inclusive of many health procedures, policies, and regulations recently integrated in school administration [
3,
4]. The merits of in-service training depend on learning transfer to the work setting (classroom) [
5]. Learning transfer was first defined as “the degree to which participants apply the knowledge, skills and attitudes acquired in training in their professional activity” [
6].
Although health guidance trainings have the potential to scale up teachers’ ability to prevent virus contamination in school [
7,
8], much less research has investigated factors that exhibit or inhibited such abilities to transfer in the classroom setting because of a lack of a better learning transfer assessment tool [
9,
10]. Past measurements of the scale of learning transfer were found to contain some flaws due to their inability to identify and evaluate all possible factors that influence learning and learning transfer both during training program itself and at the workplace [
11] Further, in an epidemic situation, the lack of a cross-culturally valid tool to quantify learning transfer predictors has been the biggest impediment to health educators, school administrators, and government official to harmonizing strategies and replicating good practices [
12]. Such tools are also needed to assess the return on investment from teachers’ training based on learning transfer at the workplace.
In the literature, many studies have highlighted the influence of contextual parameters on the measurement of learning transfer [
13,
14,
15] more specifically with regard to the administration context of the measure (e.g., source of transfer evaluation, time of transfer measurement, etc.) and parameters of the training itself (e.g., duration of training, type of skills developed, pedagogical strategies employed). According to Taylor et al. [
16], the source of transfer evaluation would influence the evaluation in such a way that it would be higher when coming from the learner (self-reported measure) as compared to an evaluation coming from another source (p e.g., immediate superior or colleagues). In the same vein, Baldwin et al. [
17] argued that there would be a significant gap between the level of transfer anticipated by learners and that achieved. In general, findings suggest that the benefit of workplace learning transfer varies from the modest 40% of training knowledge transfer six months after training [
16] to a low 10% rate of training transfer after a year [
18]. Other studies have reported that certain behavioral changes can sometimes take up to a year to appear, in particular due to constraints present in the workplace or specific characteristics of the behaviors being transferred [
19]. Thus, in addition to the characteristics of the training already recognized to promote the transfer of learning (e.g., relevance of learning objectives, relevance of training content, activities for putting learning into practice during training) [
13,
14,
15] added new characteristics relating to the current context of organizations, such as the mode of dissemination (formal vs. informal; self-learning, peer learning, or classroom learning, etc.) or even the trainer himself [
17]. Comprehensive measurement tools would be helpful; first, for the cost–benefit analysis of in-service training such as epidemic prevention training, and, second, for understanding the key factors that enhance learning transfer to actively promote those factors and, at the same time, work toward eliminating inhibitive barriers against transfer. Our study aimed to validate a workplace learning-transfer measure among Thai school teachers.
In developing countries such as Thailand, there is a keen interest in upskilling teachers using in-service training [
9] and in the evidence base for the benefit to work performance from the training [
12,
20]. In the past, the aim of in-service training was to address issues related to group heterogeneity among teachers (e.g., various development stages; unskilled, mechanical, routine, and professional stages) and to give regular updates to changes in curriculum and pedagogy [
21]. Recently, the COVID-19 pandemic has generated a significant scale of transformation in the educational system, which requires updated knowledge to be transferred in the classroom [
22]. In Thailand, however, evidence for such transfer from the workshop to the workplace has yet to be evaluated [
23,
24]. Little is known about the factors associated with the transfer of learning, both in the training setting and in the physical environment of the classroom after training.
Further, assessing the findings and analyzing the variables that have an impact on employees’ learning transfer are often complicated when the biographical characteristics of the employee are taken into consideration [
25]. For instance, even though there are no significant variances among men and women in their problem-solving capabilities, critical skills, competitive drive, motivation, learning ability, or friendliness, women are said to be more conforming and have lower achievement expectations than males [
25]. In general, when it comes to training participation, several researchers have found that females have a more positive attitude than males [
16,
26]. Thus, their perception of learning transfer may differ. In the context of a female-dominated sector like education, this study sought to validate a widely used measure of a learning transfer inventory system (LTIS) in a sample of Thai teachers and to test for measurement invariance and the latent mean difference of the instrument across gender.
While Thailand is grappling with the third wave of the COVID-19 epidemic, teachers’ lack of evidence for the workplace transfer of learning after in-service training on an epidemic control program is concerning for the Ministry of Education (MOE) in Thailand [
27]. The goal of LTIS instruments in a local school setting is to offer evidence for school principals to proactively address problems that prevent worker to transfer the trained skills, such as health code guidance for epidemic prevention. In such a context, indicators of the transfer of learning include: (1) the ability of learners to apply their learning in work settings that are more or less comparable to those provided in the context of the training (“generalization of knowledge”) [
28] and (2) the ability to apply their new learning outcomes over time, not just in the immediate aftermath of the training (maintenance over time) [
5]. To assess such indicators, according to the meta-analysis of Blume et al. [
14], factors that influence the transfer of learning can span from (1) individual variables (e.g., self-efficacy, motivation, attitudes), (2) variables relating to training (e.g., training design), and (3) the work environment (e.g., resistance to change). In addition, the perceived usefulness of the training or the value that the learner attributes to his/her participation in the training would be relevant to the transfer of learning to the actual workplace [
16]. Similarly, environmental elements, such as organizational culture [
16] and the support of peers [
29] or that provided by the immediate superior [
30], are extremely important to impacting the transfer process [
20].
Among the measures of learning transfer, the learning transfer inventory system (LTIS, Holton, et al. [
31]) is among the most widely used methods for evaluating training transfer across organizations, training types, and multiple cultural contexts. The LTIS measures training transfer in two dimensions including (1) 11 factors for training specific domain (e.g., learning readiness, motivation to transfer) and (2) 5 factors for training general domain (e.g., resistance/openness to change, performance, self-efficacy) [
32,
33]. In the current COVID-19 pandemic, the LTSI can assess the effectiveness of the measures for epidemic control implemented in the school setting after training to safely keep schools open. It comprehensively identifies and maps the evidence of training transfer both in the training session and the classroom setting and therefore establishes room for improvement. With previous cross cultural validation in the United States [
32], France [
34], Germany [
35], New Zealand [
36], Saudi Arabia [
37], and South Korea [
33] but less so in developing countries, the LTSI is a reliable diagnostic tool to conduct comparative learning transfer research across cultural boundaries. Empirically, convergent, divergent, predictive, and criterion-related validity findings have been discovered using both the original English and translated versions of the instrument in some studies, supporting the instrument’s psychometric soundness [
32,
33]. Thus, from a theoretical perspective, the LTSI validation in Thailand represents the possibility of identifying a nomological network of the learning transfer system. Some findings on one hand supported the instrument’s 16-factor structure [
29,
32,
33], and other studies have reported a 12-factor model of the LTSI [
38], calling for further studies to clarify the structure of the LTSI. There may be gendered differences in learning transfer yet to be assessed in the use of the LTSI, given the fact that in female-dominant fields such as education, gender difference has been shown to lead to difference perception in training transfer [
39]. Validation of the utilized latent trait scores may help resolve the inconsistencies in the LTSI factorial structure, and no studies examining the LTSI’s measurement invariance across gender have been conducted (Bates et al., 2012).
Goal of the Study
We validated the LTSI in the sample of Thai high school teachers. In doing so, we hypothesized that, with regard to the scores from the LTSI by teachers:
Hypothesis 1 (H1). The LTIS shows acceptable construct validity.
Hypothesis 2 (H2). The LTIS shows acceptable discriminant validity.
Hypothesis 3 (H3). The LTIS shows acceptable convergent validity.
Hypothesis 4 (H4). The LTIS yields reliable scores.
4. Discussion
With respect to the training in the specific domain, 33 items were retained for 11 factors identical to the original version. This includes (1) learner readiness, (2) personal outcomes negative, (3) personal capacity for transfer, (4) peer support, (5) supervisory/manager support, (6) supervisory/manager sanction, (7) motivation to transfer training, (8) transfer design, (9) opportunity to use learning, (10) personal outcomes: positive, and (11) personal outcomes: negative. Previous studies report similar findings [
31,
32,
33,
34,
35]. In a study done in South Korea, the authors tested the validity of the Korean-translated LTSI across 16 organizations, including education institutions. The sample data collected (753) was split for exploratory and confirmatory factors analysis. The result showed that the LTSI scale is a valid instrument measuring training in 11 specific constructs [
33]. Likewise, a study done in Jordan and Germany also resulted in retaining 33 items for 11 constructs for training in the specific domain [
35,
46].
Regarding training in the general domain, 15 items were retained and grouped into 5 factors also identical with the original version of the LTSI factors. These factors include (1) resistance/openness to change, (2) performance self-efficacy, (3) feedback/performance coaching; (4) transfer effort performance expectancy, and (5) transfer effort performance expectations. These results are in line with previous studies [
16,
47,
48,
49]. The application of the LTSI among New Zealand physical education teachers revealed results consistent with the 5 factors of training in the general domain [
36]. In a similar study done in South Africa using only an exploratory factor analysis, the five general domain factors were retained as the most-significant work environment indicators of learning transfer [
50].
Scores from the LTSI had discriminant validity in Thailand. A recent study done by Kim et al. [
33] using a procedure similar to the one adopted in this study showed a clear replication of the results obtained in the original study done by Bates, Holton, and Paul [
32] and by this study. However, due to overlap in constructs, discriminant validity is not as strong as convergent validity in self-report surveys [
51].
The comparisons among demographic participants showed that the Thais’ transfer system was different across gender. In this study, the latent mean difference analyses showed only three of sixteen scales were significantly different between male and female participants. Male respondents rated peer support, supervisor sanction, and training design higher than female respondents, but the effect was small. This gender difference could be explained from the contextual and cultural perspective [
21]. In Thailand, the formal training of teachers is mainly handled by the government. As teachers receive insufficient attention in the post-training stage, they bear responsibilities to learn from experienced peers especially in male-dominated teaching subjects like science and mathematics [
52]. In the same vein, Thai society propagates the ideology of motherhood, which restricts women’s mobility [
53]. Such restriction prevents female teachers from participating in training activities outside school premises and after teaching hours. In addition, in this gendered culture, the ideology shapes different lives for men and women by placing them in different social positions and patterns of expectation [
54]. This explains why male teachers are more concerned about supervisor sanction.