When formal training leads to informal learning among colleagues: antecedents of training transfer and knowledge sharing

To stay competitive in dynamic environments, organizations must integrate formal and informal learning. Training transfer refers to the application of knowledge and skills acquired in formal training to the workplace, whereas knowledge sharing refers to the informal dissemination of learned content to colleagues after training. Understanding the antecedents of these processes is essential for maximizing the long-term impact of training. Drawing on a two-study design, we examined antecedents of training transfer and knowledge sharing at the individual, training, and organizational clusters. Study 1 (N = 537, German-speaking employees) explored the factor structure of antecedents, while Study 2 (N = 507, U.S. employees) used a two-wave, time-lagged design to test predictive relationships in joint structural equation models. The study particularly focused on motivation and volition as underlying mechanisms linking antecedents to both training transfer and knowledge sharing. Several antecedents across the individual, training, and organizational clusters predicted motivation to transfer and/or motivation to share, although the pattern of effects differed across clusters and processes. At the individual cluster, growth orientation, training reflection, goal clarity, and knowledge sharing self-efficacy emerged as relevant predictors. At the training cluster, transfer design, training materials, content applicability, and exchange within the training group showed process-specific effects. At the organizational cluster, peer support and feedback were relevant across both processes, while supervisor support and transfer capacity showed more specific associations. Across all models, motivation was positively associated with volition, and both motivation and volition were associated with the respective post-training behavior. A central contribution lies in highlighting post-training knowledge sharing, an informal process critical for sustaining organizational learning yet rarely studied. Training transfer and knowledge sharing were predicted by overlapping antecedents, with both processes operating through motivation and volition. This shows how formal and informal learning can be shaped by common mechanisms, fostering sustainable workplace learning. The results underscore the value of integrating both perspectives to advance theory and guide interventions.

In response to rapidly evolving market conditions, societal expectations, and technological developments, organizations and employees must continuously adapt and engage in lifelong learning (Beier et al., 2025; Cascio and Montealegre, 2016; Kraiger and Ford, 2021). Enhancing employability in this dynamic environment requires employees to acquire new competencies, refine existing skills, and adjust to changing work processes (Cerasoli et al., 2018; Decius et al., 2025a). According to the World Economic Forum (2023), upskilling and reskilling will soon be necessary for approximately half of the global workforce, underscoring the urgency of developing effective learning strategies across individual, team, and organizational levels.

Formal training plays a pivotal role in supporting employee competence development and is increasingly recognized as essential (Bonekamp and Sure, 2015; Hilkenmeier et al., 2026). However, to fully leverage its potential, organizations must combine formal learning with informal learning processes tailored to individual learning paths (Blume et al., 2024; Decius et al., 2022, 2023; Decius and Decius, 2022; Poell, 2017; Richter et al., 2020). Informal learning refers to learning activities that occur outside formal training and are embedded in everyday work processes. It includes activities such as reflection, feedback, problem solving, and exchange with others, and thus represents a central part of workplace learning. In this context, knowledge sharing can be understood as one social-based aspect of informal learning, as colleagues learn from each other when newly acquired knowledge is made available and discussed in everyday work (Cerasoli et al., 2018; Tannenbaum et al., 2010). A central challenge lies in ensuring that the knowledge and skills gained during training are effectively applied in practice, a process referred to as training transfer (Baldwin and Ford, 1988). Successful training transfer contributes to organizational outcomes such as increased productivity, job satisfaction, improved performance, and profitability (Ford et al., 2018; Noe, 2017).

Yet, organizations often face time and cost constraints when providing training opportunities, which may limit employees' participation in formal training programs (OECD, 2021). This makes it even more critical that the knowledge acquired by trainees is shared with colleagues who did not attend the training. Knowledge sharing, defined as the act of making knowledge available to others within the organization (Ipe, 2003), enables informal learning through colleagues (Decius et al., 2019; Kortsch et al., 2019) and supports process improvement, problem-solving, and the development of new organizational knowledge (Ackermann et al., 2020; Fasbender and Gerpott, 2021; Ferrer-Serrano et al., 2022). At the same time, knowledge sharing may also facilitate training transfer by helping employees reflect on newly acquired content, exchange ideas about its relevance, and develop a shared understanding of how it can be applied in everyday work. Thus, knowledge sharing is relevant not only for disseminating training outcomes, but also for supporting their processing and implementation in practice. By sharing their newly acquired knowledge with colleagues, training participants enable others to learn informally from them. A contribution of this study is therefore to demonstrate how formal learning through training may lead to subsequent informal learning, for example when colleagues within the organization benefit from the participant's training experience.

Formal and informal learning should not be understood as isolated domains, but as closely interconnected processes in workplace learning. Prior research suggests that formal learning can stimulate subsequent informal learning, for example when employees continue to reflect on, discuss, and adapt newly acquired content in everyday work (Choi and Jacobs, 2011). At the same time, informal learning processes such as reflection, feedback, and exchange with colleagues are themselves central mechanisms of workplace learning and may also support the application of training content, suggesting that the relationship between formal and informal learning is not strictly one-directional (Decius et al., 2025a).

Both training transfer and knowledge sharing are key mechanisms for ensuring the sustainable impact of training initiatives (Ackermann et al., 2020; Kauffeld et al., 2025; Noe et al., 2014). To maximize these effects, it is essential to understand their antecedents and underlying mechanisms, as well as to measure them efficiently (Baldwin and Ford, 1988; Mehner et al., 2025). This paper seeks to contribute to this aim by proposing an integrated approach to assessing the processes of training transfer and knowledge sharing and their antecedents. In the present study, motivation and volition are conceptualized as central process variables that explain how antecedents are translated into training transfer and post-training knowledge sharing. The inclusion of volition adds conceptual value because the model does not only address whether employees intend to apply or share training content, but also whether they maintain and protect this intention when barriers arise during implementation (Seiberling and Kauffeld, 2017).

Accordingly, this paper makes three main contributions to the training and development literature. First, it integrates the concepts of training transfer and knowledge sharing, which are often examined in isolation, into a joint framework that emphasizes their complementary roles in ensuring the sustainable impact of training initiatives. Second, it identifies and systematizes the antecedents and underlying mechanisms of both processes, thereby advancing theoretical understanding of how learning outcomes extend beyond the individual to influence teams and organizations. Third, it introduces an integrated approach to assessing training transfer and knowledge sharing that not only provides practical guidance for designing training initiatives, but also illustrates how formal learning through training can spark subsequent informal learning when colleagues benefit from the participant's newly acquired knowledge.

Theoretical background

Training transfer and knowledge sharing

Both training transfer and knowledge sharing are pivotal processes for ensuring the sustainable impact of training initiatives (Ackermann et al., 2020; Noe, 2017). Training transfer enables employees to apply newly acquired knowledge and skills, enhancing performance, reducing errors, and boosting job satisfaction (Bonekamp and Sure, 2015; Salas et al., 2012). However, empirical findings indicate that often less than 30% of training content is actually implemented at work (Ford et al., 2018; Saks and Belcourt, 2006). To enhance training transfer, it is necessary to understand the process and the antecedents of it (Baldwin and Ford, 1988; Grossman and Salas, 2011).

Seiberling and Kauffeld (2017) proposed a model to explain the training transfer process. While building on the established concept of motivation to transfer, defined as the intention to apply newly acquired knowledge and skills in the workplace (e.g., Burke and Hutchins, 2007; Devos et al., 2007; Gegenfurtner, 2013), the model extends this perspective by incorporating volition to transfer, grounded in the Rubicon model (Achtziger and Gollwitzer, 2008). Volition to transfer refers to the ability to make intentional choices and decisions and to maintain goal-directed behavior despite barriers or resistance (Seiberling and Kauffeld, 2017). Their findings indicate that motivation and volition to transfer represent separate constructs, each functioning as an important mediator in accounting for variance in training transfer. Mehner et al. (2025) further confirmed the model empirically, showing that motivation and volition to transfer are distinct constructs and act as effective mediators of training transfer outcomes.

Beyond individual application, the sharing of newly acquired knowledge plays a critical role in maximizing the organizational impact of training (Holton et al., 2000). While a substantial body of research has examined knowledge sharing in general (Cabrera et al., 2006; Tohidinia and Mosakhani, 2010; Wen and Wang, 2022), empirical studies that focus specifically on post-training knowledge sharing remain scarce (Baharim, 2008; Butler et al., 2006). Knowledge sharing in general promotes team learning, enables ongoing process improvement, and fosters the long-term retention of knowledge within organizations (Ackermann et al., 2020; Decius et al., 2019).

As with training transfer, both motivation and volition have been identified as relevant mechanisms in the knowledge sharing process (Burmeister and Deller, 2016; Chow and Chan, 2008). In line with this, Mehner et al. (2025) provided evidence that motivation and volition substantially affect knowledge sharing behavior following training. Their model examined training transfer and knowledge sharing in combination and confirmed that both processes jointly predict positive organizational outcomes. In this study, knowledge sharing and training transfer are treated as separate processes (Mehner et al., 2025). In both processes, motivation affects volition, which in turn predicts the respective behavior. Motivation is also assumed to exert a direct influence on behavior. In the present study, this path from motivation to volition and subsequently to behavior and from motivation to behavior directly is referred to as the training transfer process and the knowledge sharing process, respectively.

Regarding post-training knowledge sharing, Mehner et al. (2025) showed that knowledge networks within organizations tend to adapt after training, with many employees sharing their newly acquired knowledge primarily with colleagues with whom they maintain close working relationships and who are perceived as likely to benefit from it. In this way, colleagues learn informally from training participants, a process that can be described as informal learning from others (Kortsch et al., 2019). At the same time, training participants themselves continue to seek and acquire knowledge informally, for instance from colleagues who are also affected by the training content or related process changes (Mehner et al., 2025). In a similar vein, Richter et al. (2020) demonstrated that employees who were more satisfied with a training program were more likely to engage in additional informal learning afterward. Thus, participants both disseminate and enrich their own knowledge through informal exchanges. This dynamic highlights the close interconnection between formal and informal learning and underlines the relevance of investigating common antecedents that predict both training transfer and knowledge sharing in the context of workplace learning.

Hypothesis 1: in both the training transfer process and the knowledge sharing process, motivation positively predicts volition, which in turn predicts the respective behavior. In addition, motivation is expected to exert a direct positive effect on behavior.

Taken together, these findings underline that training transfer and knowledge sharing represent distinct but complementary post-training processes that extend the impact of formal training through subsequent informal learning. To better understand how these processes emerge and can be supported, it is essential to examine the conditions that enable or hinder them. In line with established frameworks, the antecedents of training transfer and knowledge sharing can be organized into three clusters: individual, training-related, and organizational (Baldwin and Ford, 1988; Grossman and Salas, 2011; Witherspoon et al., 2013). The following section outlines these antecedents in greater detail and introduces the constructs investigated in this study.

Antecedents (of training transfer and knowledge sharing)

While extensive research exists on the antecedents of training transfer and its measurement (Gegenfurtner, 2009; Holton et al., 2000; Sandmeier et al., 2021), empirical research that directly examines post-training knowledge sharing and its antecedents is still limited (Baharim, 2008; Butler et al., 2006). Moreover, despite conceptual similarities between the two constructs, there is limited research investigating shared antecedents of training transfer and knowledge sharing (Hilkenmeier et al., 2026; Kauffeld et al., 2025; Mehner et al., 2025). Given that both processes can be jointly modeled (Mehner et al., 2025), it is both theoretically and practically meaningful to explore common antecedents and assess their relevance together for training transfer and knowledge sharing. Efficient measurement of these antecedents allows organizations to implement targeted adjustments, where necessary, to enhance training effectiveness. At the same time, investigating training transfer and knowledge sharing together also provides insights into the interplay of formal and informal learning in organizations. While training transfer primarily reflects the application of knowledge and skills acquired in structured learning settings, knowledge sharing captures the informal sharing of this knowledge and these skills in everyday work. Examining both processes together allows for a more comprehensive understanding of how formal training outcomes are sustained and amplified through informal learning practices.

Building on this three-cluster framework, the present study examines specific antecedents within each domain (Baldwin and Ford, 1988; Grossman and Salas, 2011; Witherspoon et al., 2013). The individual cluster refers to personal dispositions and psychological processes that influence whether and how newly acquired knowledge and skills are applied and shared. The training cluster includes features of the learning environment and instructional design that shape participants' engagement and subsequent behavior. The organizational cluster encompasses structural, social, and procedural aspects of the work environment that may either facilitate or inhibit learning transfer and knowledge sharing (Baldwin and Ford, 1988; Burke and Hutchins, 2007; Gegenfurtner et al., 2009; Renta-Davids et al., 2014). The antecedents in this study were examined in these clusters of conceptually related constructs because the aim was not only to assess each antecedent separately, but also to test whether related antecedents within each theoretical domain can be empirically distinguished from one another.

As emphasized, effective transfer and dissemination of learning outcomes are not solely dependent on individual effort. While learners may be motivated and capable of self-regulation, their success is embedded within a broader system. Without contextual alignment, transfer barriers may persist and interventions may fail to realize their full impact. Understanding the antecedents on these clusters is therefore essential for the development and evaluation of targeted support measures (Baldwin and Ford, 1988; Burke and Hutchins, 2007; Grossman and Salas, 2011; Holton et al., 2000) (Figure 1).

The present study does not aim to cover all possible antecedents of training transfer and post-training knowledge sharing. Instead, we selected a focused set of antecedents based on three criteria: first, their theoretical relevance within established transfer frameworks; second, their empirical relevance in prior reviews and meta-analytic work; and third, their conceptual relevance for both training transfer and post-training knowledge sharing as well as their practical diagnostic and intervention value for organizations. The overall structure of the antecedents was guided by established transfer frameworks that distinguish between individual or trainee characteristics, training design or intervention-related factors, and work environment characteristics (Baldwin and Ford, 1988; Burke and Hutchins, 2007; Grossman and Salas, 2011; Holton et al., 2000). Within this structure, we prioritized constructs that have repeatedly been discussed as relevant predictors in prior empirical and integrative work on training transfer, including motivational orientations, self-efficacy, perceived applicability and transfer design, supervisor and peer support, opportunities to apply training content, feedback, and transfer capacity (Blume et al., 2010; Gegenfurtner, 2009, 2013; Grossman and Salas, 2011; Holton et al., 2000).

To ensure that the selected antecedents were also meaningful for post-training knowledge sharing, we additionally drew on research on knowledge sharing antecedents. Reviews and meta-analytic work in this field emphasize individual and motivational factors, interpersonal and team-related conditions, and organizational context as central domains for explaining knowledge sharing intentions and behavior (Cabrera et al., 2006; Wang and Noe, 2010; Witherspoon et al., 2013). These perspectives supported the inclusion of constructs such as knowledge sharing self-efficacy, perceived usefulness and applicability of training content, exchange with peers, social support, feedback, and contextual opportunities and resources. Finally, qualitative evidence on post-training transfer and knowledge sharing was used to identify proximal and practically actionable facilitators that are particularly relevant after training, such as training materials, exchange within the training group, supervisor and peer support, and opportunities to apply and disseminate newly acquired knowledge (Mehner and Kauffeld, 2023; Mehner et al., 2025).

Accordingly, the selected antecedents represent a theory-based, empirically informed, and practice-oriented subset rather than an exhaustive catalog of all possible predictors. Some antecedents discussed in broader transfer frameworks, such as cognitive ability or general transfer climate, were not included in their broader form. Cognitive ability was not prioritized because the present study focused on proximal and modifiable antecedents that can be addressed through training design, diagnostic feedback, and workplace support. Similarly, general transfer climate was represented through more specific and actionable contextual indicators, such as supervisor support, peer support, feedback, opportunity to apply, and transfer capacity.

In line with the process model outlined above, this study focuses on the direct effects of the antecedents on motivation to transfer and motivation to share. This modeling decision is grounded in the assumed temporal and process logic of post-training behavior. Drawing on the Rubicon model of action phases, motivation reflects the predecisional and preactional orientation toward forming an intention, whereas volition becomes particularly relevant once an intention has been formed and must be maintained and protected during implementation despite barriers (Achtziger and Gollwitzer, 2008). Applied to training transfer, Seiberling and Kauffeld (2017) conceptualized motivation to transfer and volition to transfer as distinct but sequential process variables, with motivation preceding volitional regulation and subsequent transfer behavior. In this logic, antecedents such as self-efficacy, perceived applicability, support, feedback, or opportunities to apply training content are assumed to first influence whether employees are motivated to use or share what they have learned. This motivation then provides the basis for volitional regulation, which supports the enactment of the intended behavior when obstacles arise.

Accordingly, motivation is conceptualized here as the most proximal starting point through which antecedents are translated into later volitional regulation and post-training behavior. Although direct effects of antecedents on volition and the behavioral outcomes are theoretically possible, these additional paths were not specified in the following hypothesis set and were not included in the main models. This decision reflects both the theoretical process logic of the motivation–volition–behavior sequence and the aim to test a parsimonious model focused on the hypothesized mechanism from antecedents to motivation and subsequently to volition and behavior.

Individual-cluster antecedents

Individual antecedents comprise a range of psychological dispositions and self-regulatory mechanisms that influence whether and how employees apply and disseminate newly acquired knowledge following a training intervention (Gegenfurtner et al., 2009)

Growth orientation reflects a proactive orientation in which individuals approach training with the intention to challenge themselves, expand competencies, and pursue personal and professional growth (Colquitt and Noe, 2000; Dweck and Yeager, 2019). Rooted in a promotion focus (Higgins, 1997), this orientation fosters openness to new experiences and enhances motivation to apply training content in practice. Empirical findings indicate that developmental goal framing is positively associated with training transfer, as learners are more likely to experiment with and integrate newly acquired skills into their work routines (Van Yperen et al., 2011). In addition, growth orientation has been linked to greater knowledge sharing, since individuals with a growth orientation are more motivated to exchange knowledge, collaborate, and contribute to collective learning (Cerne et al., 2014).

Deficit orientation describes an orientation in which individuals participate in training primarily to maintain their current job performance and to prevent mistakes or performance decline (Higgins, 1997; Elliot and McGregor, 2001). While this prevention-focused orientation emphasizes safeguarding the status quo rather than pursuing growth (Higgins, 1997), it can still contribute positively to training transfer. Specifically, deficit orientation provides motivation to engage with training content in order to avoid negative outcomes, which supports the application of acquired knowledge and skills in practice. However, compared to growth orientation, its positive effects on transfer tend to be weaker, as the focus on error prevention limits exploratory engagement and broader skill development (Van Yperen et al., 2011). In the context of knowledge sharing, deficit orientation may provide a basic level of motivation to exchange knowledge when sharing serves a risk-reduction function. By discussing newly learned content with colleagues, trainees may clarify their own understanding, receive corrective feedback, and align the application of training content with others, thereby reducing the likelihood of implementation mistakes. However, because the underlying aim is primarily to avoid errors or performance decline rather than to explore new ideas, deficit orientation is less likely to foster proactive or collaborative sharing behaviors beyond what is necessary for safe implementation (Higgins, 1997; Elliot and McGregor, 2001; Cerne et al., 2014).

Transfer self-efficacy refers to learners' confidence in their ability to successfully apply knowledge and skills acquired during training to their everyday work situations (Holton et al., 2000; Samim, 2018). High levels of transfer self-efficacy are a well-established predictor of training transfer, as trainees who believe in their capacity to implement new skills are more likely to persist in applying them and to overcome barriers in the work environment (Chiaburu and Marinova, 2005). Beyond its role in training transfer, self-efficacy has also been shown to positively influence knowledge sharing, since employees with stronger self-beliefs are more inclined to contribute their expertise and exchange knowledge with colleagues (Cabrera et al., 2006). Accordingly, transfer self-efficacy may not only strengthen the application of training content but also foster motivation to share knowledge after training.

Knowledge sharing self-efficacy refers to individuals' confidence in their ability to effectively communicate and share their knowledge with others in the workplace (Bock and Kim, 2002; Kankanhalli et al., 2005). High levels of knowledge sharing self-efficacy have been shown to positively predict knowledge sharing behavior, as employees who believe in their capacity to articulate and exchange knowledge are more willing to engage in such practices (Cabrera et al., 2006). Moreover, self-efficacy in knowledge sharing may also reinforce training transfer, since trainees who feel capable of explaining and disseminating newly acquired knowledge are likely to engage more deeply with training content and strengthen its application in practice.

Training reflection refers to the cognitive process through which learners revisit, analyze, and interpret training content and their learning experiences (Decius et al., 2019; Sparr et al., 2017). This process involves reexamining what was learned, identifying gaps in understanding, integrating new information with prior knowledge, and planning how to apply insights in future contexts. Reflective practice is considered a critical mechanism for promoting deeper learning and enhancing the transfer of training to real-world application (Riesen and Whitver, 2023). Moreover, engaging in reflection by consciously considering and articulating one's knowledge can facilitate knowledge sharing, as it encourages the externalization of insights that would otherwise remain tacit; reflective cognitive activities have been empirically associated with increased knowledge dissemination in organizational contexts (Sun et al., 2022). Therefore, reflection on training content may not only strengthen training transfer but also foster motivation to share knowledge after training.

Goal clarity, which is conceptually anchored in the learner readiness construct of the Learning Transfer System Inventory (Holton et al., 2000), refers to the extent to which trainees have a clear understanding of the objectives, expectations, and intended outcomes of a training program. It captures the degree to which participants know what they are expected to learn and how the acquired knowledge and skills are intended to be applied in their work context. Clear articulation of training goals enhances training transfer, as participants who understand expected outcomes are better able to align their efforts and sustain behavioral changes on the job (Johnson et al., 2012). Goal clarity could also enhance knowledge sharing by fostering a shared understanding of objectives, which in turn promotes the exchange of information and expertise among team members (Van der Hoek et al., 2018).

These individual-cluster antecedents represent essential psychological mechanisms that jointly influence whether employees apply and disseminate knowledge after training. Previous research has demonstrated that motivational orientations (e.g., developmental and deficit orientation), efficacy beliefs (e.g., transfer and knowledge sharing self-efficacy), reflective practices, and goal clarity each contribute to training transfer and knowledge sharing in distinct yet complementary ways (Chiaburu and Marinova, 2005; Johnson et al., 2012; Riesen and Whitver, 2023; Sun et al., 2022). Building on this evidence, the present study assumes that these antecedents can be empirically distinguished as separate constructs and that each predicts training transfer and knowledge sharing processes.

Hypothesis 2.1: The individual-cluster antecedents of training transfer and knowledge sharing are represented by empirically distinct and internally consistent constructs that are each adequately measured by their respective items.

Hypothesis 2.2: Each individual-cluster antecedent positively predicts motivation to transfer and motivation to share.

Training-cluster antecedents

Training-cluster antecedents refer to characteristics of the formal learning setting that can facilitate or hinder the later application and dissemination of training content. They capture the extent to which the training itself provides relevant content, transfer-oriented design, supportive materials, and opportunities for exchange that foster post-training motivation and behavior (Baldwin and Ford, 1988; Grossman and Salas, 2011).

Content applicability refers to the perceived relevance and applicability of the training content to the trainees' job tasks, it captures whether participants consider the training material as useful and applicable to their day-to-day work (Holton et al., 2000). When trainees perceive the training content as applicable to their job (i.e., high content applicability), they are more likely to be motivated to implement the acquired knowledge and skills in their work environment, which positively affects training transfer (Gegenfurtner et al., 2009; Holton et al., 2000). Training content that is seen as highly applicable may also enhance knowledge sharing behaviors, as employees are more likely to share knowledge they perceive as useful and relevant to others in their work context (Kim and Lee, 2006; Lin, 2007).

Transfer design refers to the degree to which a training program has been intentionally structured to promote the application of learning to the job context. It encompasses the instructional strategies, sequencing, practice opportunities, and alignment with transfer goals embedded in the training itself (Holton et al., 2000). Transfer design has consistently been shown to enhance training transfer (Holton et al., 2000; Renta-Davids et al., 2014; Santana-Domínguez, 2022). Although there are fewer empirical studies specifically linking transfer design to knowledge sharing, elements of transfer-design in training, such as clear expectations, relevant content, active practice, and explicit transfer planning align with key mechanisms in knowledge sharing and organizational knowledge transfer, suggesting that well-structured training design can indirectly facilitate post-training knowledge dissemination within organizations (Argote, 2022). Transfer design is also recognized for its integrative function, serving as a mechanism to connect formal, informal, and self-regulated learning approaches (Kauffeld et al., 2025).

Training materials refer to the extent, quality, and degree of structure with which training-related documents are provided to participants (Mehner and Kauffeld, 2023; Nielsen et al., 2023). Well-structured training materials can enhance training transfer and can be a critical resource that not only support training transfer but also facilitate post-training knowledge sharing, as they provide tangible, reusable references that reinforce learning and promote the dissemination of acquired knowledge (Mehner and Kauffeld, 2023; Nielsen et al., 2023).

Exchange within the training group refers to the extent to which interactions with fellow participants during the training were perceived as enriching (Mehner and Kauffeld, 2023; Salamon et al., 2023). Peer group interaction has been conceptually linked to enhanced training effectiveness, particularly when mediated by transfer motivation and a supportive transfer climate (Elembiassery, 2016). In a qualitative interview study, this factor emerged as a frequent facilitator of both training transfer and post-training knowledge sharing. Participants reported that these interactions not only supported immediate learning but also often extended beyond the training itself, fostering continued exchange on training-related topics (Mehner and Kauffeld, 2023).

These training-cluster antecedents represent central characteristics of the learning environment that shape whether employees implement and share newly acquired knowledge after training. Prior studies have shown that content applicability, transfer design, training materials, and group exchange each contribute to training transfer and knowledge sharing in distinct but complementary ways (Gegenfurtner et al., 2009; Holton et al., 2000; Kim and Lee, 2006; Mehner and Kauffeld, 2023). Building on this evidence, the present study assumes that these antecedents can be empirically distinguished as separate constructs and that each predicts the training transfer and knowledge sharing processes.

Hypothesis 3.1: The training-cluster antecedents of training transfer and knowledge sharing are represented by empirically distinct and internally consistent constructs that are each adequately measured by their respective items.

Hypothesis 3.2: Each training-cluster antecedent positively predicts motivation to transfer and motivation to share.

Organizational-cluster antecedents

Organizational-cluster antecedents refer to features of the work environment that shape whether newly acquired knowledge can be applied and shared after training. They include social support, feedback, opportunities, and resources that may either enable or constrain the continuation of learning in everyday work (Baldwin and Ford, 1988; Witherspoon et al., 2013).

Supervisor support refers to the degree to which leaders actively promote and facilitate employees' learning and application of newly acquired knowledge and skills in the workplace (Hughes et al., 2020; Reinhold et al., 2018). This can include actions such as engaging in conversations about recently acquired knowledge and skills and signaling interest in employees' development, and reinforcing learning transfer (Seiberling and Kauffeld, 2017; Lim and Johnson, 2002). Moreover, supervisor support has been shown to positively influence knowledge sharing behaviors (Cabrera et al., 2006; Massenberg and Kauffeld, 2015; Riege, 2005). Supervisors also contribute to a supportive knowledge-sharing climate by removing obstacles, providing access to resources, allocating time, encouraging participation, and serving as behavioral role models (Buch et al., 2015; Ferrer-Serrano et al., 2022).

Peer support refers to the supportive behaviors among colleagues that involve offering assistance, sharing experiences, and encouraging each other in the application of newly acquired knowledge and skills (Hughes et al., 2020; Reinhold et al., 2018). It has been identified as an important factor influencing training transfer, as demonstrated in numerous empirical studies (e.g., Baldwin and Ford, 1988; Bates et al., 2007; Gegenfurtner et al., 2009; Holton et al., 2000; Massenberg et al., 2015; Sarfraz et al., 2020). Through mutual support, peers can strengthen confidence in using newly acquired knowledge and skills and create opportunities for exchanging practical insights, thereby fostering both transfer and knowledge sharing (Ferrer-Serrano et al., 2022; Wang and Noe, 2010).

Opportunity to apply refers to the extent to which trainees are provided with situational possibilities to apply newly acquired knowledge and skills in their job context after training. It captures the degree to which the work environment gives opportunities for the practical implementation (Blume et al., 2010). Having sufficient opportunity to apply newly acquired knowledge and skills is positively associated with transfer motivation (Gegenfurtner, 2009). Opportunity to apply also foster knowledge sharing and knowledge management processes. When trainees are given sufficient opportunities to apply newly acquired knowledge and skills, they generate explicit outputs and practical experiences that can be more easily communicated to others, thereby supporting motivation for knowledge dissemination (Argote and Ingram, 2000).

Feedback refers to the extent to which trainees receive evaluative information regarding their application of training through the work itself, formal reviews or informal cues (Holton et al., 2000). Such feedback may stem directly from task outcomes, performance indicators, or everyday work experiences, and can also be complemented by more structured evaluations (Humphrey et al., 2007). Empirical research supports feedback as a factor that enhances training transfer: for example, increased feedback specificity has been shown to improve transfer outcomes by reinforcing relevant behavior and guiding appropriate adjustments (Goodman, 2011). In addition, feedback processes, whether derived from the task environment or interpersonal interactions, have been found to positively influence knowledge sharing behaviors (Cabrera et al., 2006; Ferrer-Serrano et al., 2022; Riege, 2005).

Transfer capacity refers to the extent to which trainees are provided with sufficient time and resources to transfer newly acquired knowledge and skills into the workplace (Holton et al., 2000). Personal transfer capacity has been shown to significantly enhance the effectiveness of training transfer (Kirwan and Birchall, 2006). From a knowledge management perspective, these same resources are essential for knowledge sharing, since lack of time and resources is among the most frequently cited barriers to exchanging knowledge in organizations (Riege, 2005). Thus, personal transfer capacity may not only enable the application of learning outcomes but also facilitate their dissemination within the workplace.

These organizational-cluster antecedents represent critical contextual conditions that jointly shape whether employees apply and disseminate training outcomes in the workplace. Prior research has shown that supervisor support, peer support, opportunity to apply, feedback, and transfer capacity each contribute to training transfer and knowledge sharing in distinct yet complementary ways (Baldwin and Ford, 1988; Blume et al., 2010; Ferrer-Serrano et al., 2022; Riege, 2005). Building on this evidence, the present study assumes that these antecedents can be empirically distinguished as separate constructs and that each predicts the training transfer and knowledge sharing processes.

Hypothesis 4.1: The organizational-cluster antecedents of training transfer and knowledge sharing are represented by empirically distinct and internally consistent constructs that are each adequately measured by their respective items.

Hypothesis 4.2: Each organizational-cluster antecedent positively predicts motivation to transfer and motivation to share.

To investigate the antecedents of training transfer and knowledge sharing, two consecutive studies were conducted. Study 1 served to explore the factor structure of the antecedent measures and reduce the item pool, whereas Study 2 was used to validate the revised scales and test the predictive relationships in a two-wave design. Further details regarding samples and procedures are provided in the following sections.

Methods study 1

Study design, procedure, and analysis

Prior to the first study, an initial item pool was developed based on a comprehensive literature review of existing scales and items for assessing the antecedents and outcomes of this study (e.g., Aguinis and Kraiger, 2009; Antunes and Nascimento, 2018; Banerjee et al., 2017; Bell et al., 2017; Bhatti et al., 2013; Böhn and Deutscher, 2021; Chauhan et al., 2017; Chatterjee et al., 2018; Friedman and Ronen, 2015; Gegenfurtner, 2009; Govaerts et al., 2018; Holton et al., 2000; Hurt, 2016; Khan et al., 2015; Konradt et al., 2008; Massenberg et al., 2015, 2017; Mehner and Kauffeld, 2023; Mehner et al., 2025; Sandmeier et al., 2021; Schindler and Burkholder, 2016; Seiberling and Kauffeld, 2017; Twase et al., 2021). The antecedents were selected based on their theoretical relevance within established transfer frameworks, their conceptual relevance for both training transfer and post-training knowledge sharing, and their practical diagnostic and intervention value.

Details on the sources used to develop the individual scales are provided in Section 3.3 (“Measures”). For the constructs Feedback and Transfer volition, existing validated scales were applied without modification (Feedback: Schulte et al., 2021; Transfer volition: Seiberling and Kauffeld, 2017).

Based on this literature, items were formulated and subsequently discussed in a pre-workshop with seven experts in work and organizational psychology, three of whom held a formal certification as workplace trainers. Following this, a cognitive pre-test was conducted with three individuals from the targeted group to evaluate and ensure the clarity and appropriateness of item wording.

Initially, five to six items per scale were developed and tested. Study 1 was then carried out in a German-speaking context with a sample of 541 participants.

All analyses were conducted using Mplus 8.4 (Muthén & Muthén, Los Angeles, CA, United States). To establish the factor structure of the three proposed clusters—individual, training, and organizational—we ran exploratory structural equation modeling (ESEM) models with maximum likelihood estimation and robust standard errors (MLR). These ESEMs included only the items of the 15 antecedent constructs and were estimated separately for the three antecedent clusters (individual, training-related, and organizational). Motivation, volition, training transfer, and knowledge sharing were not included in these models. Model fit was evaluated based on multiple criteria: the chi-square test (χ 2), the Root Mean Square Error of Approximation (RMSEA; values < 0.08 considered acceptable; MacCallum et al., 1996), the Comparative Fit Index (CFI; values >0.90 considered acceptable), the Tucker–Lewis Index (TLI; values >0.90 considered acceptable), and the Standardized Root Mean Square Residual (SRMR; values < 0.08 considered good; Hu and Bentler, 1999). With respect to factor loadings, we followed the rule of thumb that items should load at least 0.50 on their intended factor and should not exhibit cross-loadings above 0.30 on any other factor (Cudeck, 2000; Tinsley and Tinsley, 1987). Based on the results, the item pool was reduced to three items per scale. Only for the Goal Clarity scale, it was not possible to retain three suitable items; therefore, one item that loaded at 0.388 in the analysis was slightly adapted.

In addition, we examined the core motivation–volition–behavior paths proposed in Hypothesis 1 for both the training transfer and knowledge sharing processes. These analyses were conducted in Mplus 8.4 using the MLR estimator. Given the cross-sectional design of Study 1, these analyses were considered preliminary and are reported in the Supplementary materials.

Participants

In the first validation phase, data were gathered from a German-speaking sample recruited via the professional panel provider Bilendi. To ensure relevance for the study context, participants had to meet specific inclusion criteria: they needed to be employed for a minimum of 20 h per week and must have completed a job-related training program within the past 2 years. These criteria were assessed at the beginning of the survey, and respondents who did not qualify were excluded from participation.

Of the 541 individuals who met the inclusion criteria, four were excluded from the final dataset due to missing responses on key outcome measures related to training transfer and knowledge sharing. This resulted in a final sample of 537 participants, with a mean age of 46.76 years (SD = 12.35). The sample consisted of 33.30% women, 66.00% men, and 0.70% individuals identifying as non-binary. On average, participants reported working 38.57 hours per week (SD = 8.21), and 46.30% held a university degree.

Organizational size was classified in line with the categories proposed by Sylvest et al. (2018): the majority of respondents were employed in large organizations with more than 250 employees (51.5%), while 22.80% worked in medium-sized companies (up to 250 employees), 14.90% in small enterprises (fewer than 50 employees), and 10.8% in micro-enterprises (fewer than 10 employees). Participants represented a diverse range of professional sectors, with the most common being humanities and social sciences (22.00%), transport, logistics, protection and security (12.40%), and raw materials, production, and manufacturing (9.30%).

On average, the most recent training program attended by participants had taken place 11.04 months prior to the survey (SD = 7.81) and lasted approximately 76.96 hours (SD = 128.43). Of these trainings, 60% were externally provided, while 40% were delivered internally, either by in-house staff or external trainers. The reported training formats primarily included professional development courses (60.30%) and interdisciplinary programs (39.70%). Regarding delivery mode, the majority of participants had attended classroom-based training (62.51%), while 46.82% reported having participated in virtual training sessions. Because participants referred to a training completed several months earlier, the present design captures retrospectively perceived antecedents and sustained post-training outcomes rather than immediate post-training reactions.

Methods study 2

Study design, procedure, and analysis

The reduced item set from Study 1 was translated into English, reviewed and optimized in a focus group, and further refined through a cognitive pre-test with two individuals from the target population. Data collection was embedded in a larger longitudinal project conducted via Prolific. The design included five measurement waves, each administered at 1-month intervals, allowing for the systematic investigation of relationships across time. In Study 2, all antecedent variables were assessed in one wave, whereas motivation, volition, training transfer, and knowledge sharing were assessed together in the subsequent wave. This design allowed us to temporally separate the assessment of antecedents from the subsequent assessment of motivation, volition, training transfer, and knowledge sharing, thereby strengthening the predictive analyses of antecedents.

An a priori power analysis was conducted during study planning. The results suggested that a larger sample would have been required to test additional direct effects of the antecedents on volition and the behavioral outcomes with sufficient statistical power. For this reason, these paths were not included in the present models, and no corresponding hypotheses were specified.

All analyses were conducted using Mplus 8.4. To test Hypotheses 2.1, 3.1, and 4.1, which proposed that the antecedents within each cluster are empirically distinct and internally consistent constructs, Confirmatory Factor Analyses (CFAs) were carried out using the WLSMV estimator. To examine the factorial validity of the three proposed clusters—individual, training-related, and organizational—separate CFAs were specified for each cluster. All latent factors were allowed to correlate freely, consistent with Mplus default specifications. To further assess discriminant validity among the antecedents within each cluster, we applied the Fornell–Larcker criterion (Fornell and Larcker, 1981). For each latent construct, the average variance extracted (AVE) was calculated based on the standardized factor loadings from the CFA models. We interpreted the Fornell–Larcker criterion as providing evidence for discriminant validity when the AVE of each construct exceeded its squared correlations with all other constructs within the same antecedent cluster. Given the conceptual relatedness of the antecedents within each cluster, these results were interpreted together with the CFA model fit, factor loadings, and latent factor correlations.

To test Hypotheses 2.2, 3.2, and 4.2, which proposed that the antecedents significantly predict motivation to transfer and motivation to share, latent mediation models were estimated using a two-step estimation strategy. First, the models were estimated using the WLSMV estimator. Because all indicators were assessed using five-point Likert-type response scales, we treated the item responses as ordered categorical. The WLSMV estimator was therefore chosen because it is appropriate for modeling ordinal indicators and provides robust standard errors and mean- and variance-adjusted test statistics. Second, the models were estimated using the ML estimator with bootstrapping to obtain bias-corrected bootstrap confidence intervals for the direct and indirect effects, which are particularly informative for evaluating mediation effects. This combination allowed us to retain an ordinal measurement approach for the primary model estimation while also providing bootstrap-based inference for the indirect effects. We estimated separate correlated-factor CFAs for each antecedent cluster. This approach was chosen because the hypotheses concerned the empirical distinctiveness of conceptually related antecedents within each cluster rather than the evaluation of each construct in isolation. We inspected the 95% confidence intervals based on bias-corrected bootstrap standard errors (using the Mplus command “cinterval (bcbootstrap)”) to infer statistical significance of the direct and indirect effects. Because the hypothesized models assume that antecedents affect post-training behavior via motivation and volition, indirect effects were also estimated and are reported in the Supplementary materials; however, these analyses are supplementary to the main hypothesis tests.

Participants

In the second validation phase, data were collected from a total of 513 participants, five individuals were excluded due to excessive missing data, resulting in a final sample of 507 participants. The average age was 43.49 years (SD = 10.54), and 51.07% identified as male.

In terms of educational background, 18% of participants reported having a high school diploma or equivalent, 47% held a bachelor's degree, 25% a master's degree, 6% a doctoral degree, and 4% reported other qualifications. Most participants were employed full-time, with 77% working between 40 h and 49 h per week. Additionally, 13% worked between 50 h and 59 h, 6% worked more than 60 h per week, and 4% reported working fewer than 20 h. A substantial proportion of the sample (65.43%) held a leadership or managerial position.

Participants were employed across a wide range of occupational sectors. The largest share worked in the service sector (32%), followed by those who selected “Other” and provided open-ended responses (21%). Furthermore, 20% were employed in industry, 19% in public service, 8% in trade and commerce, and 1% in crafts.

The sample was predominantly composed of U.S. nationals (98%), while the remaining 2% reported other nationalities, including German, Chinese, Vietnamese, Filipino, Polish, and Nigerian.

On average, participants had completed their most recent job-related training 7.02 months prior to the first survey wave (SD = 19.66).

Measures

All items were rated on a five-point Likert scale ranging from 1 (strongly disagree) to 5 (strongly agree). Each scale was measured with three items. Table 1 reports the means, standard deviations, internal consistencies, and intercorrelations for all scales. Consistent with recommended benchmarks, all scales demonstrated good to excellent reliability (Nunnally and Bernstein, 1994).

Motivation to Transfer items are based on the GLTSI (Kauffeld et al., 2008; Gegenfurtner, 2009; Friedman and Ronen, 2015; Sandmeier et al., 2021). Example item: I made an effort to use the training content at work.

Volition to Transfer was measured with the original three items from Seiberling and Kauffeld (2017). Example item: I was determined to apply what I had learned, even if I had to overcome resistance.

Training Transfer was assessed with items adapted from the Q4TE questionnaire by Grohmann and Kauffeld (2013) and the MEI questionnaire by Kauffeld et al. (2009) and used before by Mehner et al. (2025). Example item: I am very successful in applying the acquired training content in my daily work.

Motivation to Share was created by rephrasing the items from motivation to transfer so that they capture knowledge sharing and used before by Mehner et al. (2025). Example item: I was motivated to share the training content with my colleagues.

Volition to Share was assessed with items reformulated from volition to transfer by Seiberling and Kauffeld (2017) to measure knowledge sharing and used before by Mehner et al. (2025). Example item: I was determined to share what I had learned with colleagues, even if I had to overcome resistance.

Knowledge Sharing was measured by adapting the training transfer items to the context of knowledge exchange, while also considering existing knowledge sharing measures (e.g., Tohidinia and Mosakhani, 2010; Wilkesmann et al., 2009) and used before by Mehner et al. (2025). Example item: I shared a lot of the training content with my colleagues.

Deficit Orientation was based on Antunes and Nascimento (2018), Böhm and Jonas (2016), Gegenfurtner (2009), Konradt et al. (2008), and Tai (2006). Example item: I chose to participate in professional training in order to avoid making mistakes with new tasks.

Growth Orientation was based on Antunes and Nascimento (2018), Böhm and Jonas (2016), Gegenfurtner (2009), Konradt et al. (2008), and Tai (2006). Example item: I chose to participate in professional training in order to continually develop myself personally.

Transfer Self-Efficacy was developed drawing on Aguinis and Kraiger (2009), Bell et al. (2017), Böhn and Deutscher (2021), Gegenfurtner (2009), and Govaerts et al. (2018). Example item: After the training, I knew exactly that I could meet the necessary requirements and apply the training content.

Knowledge Sharing Self-Efficacy was derived from the transfer self-efficacy scale with a focus on knowledge sharing, based on Aguinis and Kraiger (2009), Bell et al. (2017), Böhn and Deutscher (2021), Gegenfurtner (2009), and Govaerts et al. (2018). Example item: After the training, I knew exactly that I could meet the necessary requirements and pass on the training content to colleagues.

Training Reflection was assessed using adapted items that were derived from the scales on Anticipatory and Subsequent Reflection by Decius and Schaper (2019). Example item: After the training, I thought about how I could best apply my new knowledge at my workplace.

Goal Clarity was based on Antunes and Nascimento (2018), Bhatti et al. (2013), Locke and Latham (2002), and Holton et al. (2000). Example item: I clearly understood what the training would be about before attending.

Content Applicability was based on Antunes and Nascimento (2018), Holton et al. (2000), and Twase et al. (2021). Example item: The examples from the training closely resembled situations from my daily work life.

Transfer Design was based on Chauhan et al. (2017), Gegenfurtner (2009), Holton et al. (2000), and Sandmeier et al. (2021). Example item: The exercises in the training prepared me well for applying the content in my work.

Training Materials was based on Gegenfurtner (2009), Holton et al. (2000), Mehner and Kauffeld (2023), and Sandmeier et al. (2021), with Mehner and Kauffeld (2023) serving as the primary source. Example item: The materials I received during the training contained all the important information.

Exchange within Training Group was based on Gegenfurtner (2009), Holton et al. (2000), Mehner and Kauffeld (2023), and Sandmeier et al. (2021), with Mehner and Kauffeld (2023) serving as the primary source. Example item: In the training group, we frequently discussed the content of the training.

Supervisor Support was based on Antunes and Nascimento (2018), Bhatti et al. (2013), Böhn and Deutscher (2021), Govaerts et al. (2018), Holton et al. (2000), Hurt (2016), Massenberg et al. (2015), Sandmeier et al. (2021), Schindler and Burkholder (2016), and Seiberling and Kauffeld (2017). Example item: My supervisor supports me in applying the training content at work.

Peer Support was based on Antunes and Nascimento (2018), Bhatti et al. (2013), Gegenfurtner (2009), Holton et al. (2000), Khan et al. (2015), Massenberg et al. (2015), and Schindler and Burkholder (2016). Example item: My colleagues do everything they can to ensure that I can apply the training content.

Opportunity to Apply was based on Aguinis and Kraiger (2009), Holton et al. (2000), and Massenberg et al. (2017). Example item: I had many opportunities at work to apply the training content.

Feedback was assessed with the original REA items from Schulte et al. (2021). Example item: I regularly receive feedback on my work results.

Transfer Capacity was based on Aguinis and Kraiger (2009), Antunes and Nascimento (2018), Holton et al. (2000), and Massenberg et al. (2017). Example item: My workload allowed me to apply the training content.

Hypothesis 1 stated that in both the training transfer process and the knowledge sharing process, motivation positively predicts volition, which in turn predicts the respective behavior. In addition, motivation was expected to exert a direct positive effect on behavior. This hypothesis was supported: across all three models estimated with the data from Study 2 for the individual-, training-, and organizational-cluster antecedents, the assumed paths were consistently confirmed. Motivation significantly predicted volition, volition significantly predicted the respective behavior, and motivation also showed a direct positive effect on behavior. In addition, preliminary analyses of the core process paths in Study 1 yielded the same pattern of results for both the training transfer and the knowledge sharing process; detailed results are reported in Supplementary Table 13. The results from Study 2 are illustrated in Figures 2–4.

Individual-cluster results

In Study 1, we conducted an exploratory structural equation model (ESEM) to examine the items of the antecedents at the individual cluster. The complete set of factor loadings for all items per scale, which formed the basis for the selection and exclusion of items, is reported in Supplementary Table 1. The model with the selected three items per scale demonstrated good fit to the data, χ 2 (60) = 104.12 ***, CFI = 0.990, TLI = 0.975, RMSEA = 0.033, SRMR = 0.012. Standardized factor loadings on the intended factors ranged from 0.537 to 0.850, and no item exhibited cross-loadings above 0.30 on any other factor.

In Study 2, the hypothesized six-factor measurement model showed acceptable incremental and residual fit, χ 2 (120) = 721.73 ***, CFI = 0.96, TLI = 0.95, SRMR = 0.04, but the RMSEA was elevated (RMSEA = 0.10). All items loaded significantly on their intended latent factors (standardized λ = 0.44–0.93, see Table 2), with most loadings being substantial. Internal consistencies (Cronbach's α) were satisfactory (see Table 1). Latent factor correlations within the CFA measurement model were positive and of moderate to very high magnitude (0.41–0.93), indicating substantial conceptual overlap among some constructs. Nevertheless, the separate factor structure, significant factor loadings, and the Fornell–Larcker results provided evidence that the constructs were empirically distinguishable. The corresponding Fornell–Larcker results are reported in Supplementary Table 2. Thus, Hypothesis 2.1 was supported, although discriminant validity should be interpreted in light of the high latent correlations among some antecedents.

The structural model formulated in Hypothesis 2.2, representing the individual antecedents, showed a good fit to the data, χ 2 (552) = 1586.46 ***, CFI = 0.99, TLI = 0.98, RMSEA = 0.06, SRMR = 0.04. The significant paths are displayed in Figure 2 (dotted lines show allowed correlations in the model), while all direct effects are reported in Supplementary Table 3, as well as the indirect effects in Supplementary Table 4.

Our results indicate that deficit orientation, growth orientation, goal clarity, and training reflection were positively associated with motivation to transfer, whereas transfer self-efficacy and knowledge sharing self-efficacy did not significantly predict motivation to transfer. Motivation to share was positively predicted by growth orientation, goal clarity, training reflection, and knowledge sharing self-efficacy; deficit orientation was not significant, whereas transfer self-efficacy showed a significant negative association. Therefore, Hypothesis 2.2 was partly supported.

Training-cluster results

In Study 1, we applied an exploratory structural equation model (ESEM) to evaluate the items of the antecedents at the training cluster. Supplementary Table 5 presents the full set of factor loadings for all items per scale, which served as the basis for item selection and exclusion. The model retaining three items per scale showed good fit to the data, χ 2 (24) = 39.44 *, CFI = 0.996, TLI = 0.988, RMSEA = 0.031, SRMR = 0.008. Standardized loadings on the intended factors ranged from 0.623 to 0.902, and no item displayed cross-loadings > 0.30 on alternative factors.

In Study 2, the proposed four-factor model for the training-related antecedents provided an excellent representation of the data, χ 2 (48) = 129.06 ***, CFI = 0.997, TLI = 0.996, RMSEA = 0.06, SRMR = 0.01. All indicators showed high and significant loadings on their designated latent constructs (standardized λ = 0.82–0.98; see Table 3). Internal consistencies (Cronbach's α) were satisfactory (see Table 1). Latent factor correlations within the CFA measurement model indicated that the training-related antecedents were substantially related, reflecting their shared focus on the training context. Despite this conceptual overlap, the separate factor structure, significant factor loadings, and the Fornell–Larcker results provided evidence that the constructs were empirically distinguishable. The corresponding Fornell–Larcker results are reported in Supplementary Table 6. Thus, Hypothesis 3.1 was supported, although discriminant validity should be interpreted cautiously given the conceptual relatedness of the training-related antecedents.

The structural model formulated in Hypothesis 3.2 showed an excellent fit to the data, χ 2 (376) = 964.24 ***, CFI = 0.99, TLI = 0.99, RMSEA = 0.06, SRMR = 0.03. The significant paths are displayed in Figure 3 (dotted lines show allowed correlations in the model), while all direct effects are reported in Supplementary Table 7, as well as the indirect effects in Supplementary Table 8.

The results indicate that transfer design, training materials, and exchange within the training group positively associated with motivation to transfer, whereas content applicability did not significantly predict motivation to transfer. Motivation to share was positively predicted by content applicability, training materials, and exchange within the training group, whereas transfer design was not significant. Therefore, Hypothesis 3.2 was partly supported.

Organizational-cluster results

In Study 1, we conducted an exploratory structural equation model (ESEM) to examine the items of the antecedents at the organizational cluster. The complete set of factor loadings for all items per scale, which formed the basis for the selection and exclusion of items, is reported in Supplementary Table 9. The model with the selected three items per scale demonstrated excellent fit to the data, χ 2 (40) = 51.12, p = 0.112, CFI = 0.998, TLI = 0.994, RMSEA = 0.02, SRMR = 0.008. Standardized factor loadings on the intended factors ranged from 0.700 to 0.909, and no item exhibited cross-loadings above 0.30 on any other factor.

In Study 2, the five-factor model representing the organizational antecedents demonstrated a good overall fit, χ 2 (80) = 487.57 ***, CFI = 0.987, TLI = 0.983, RMSEA = 0.04, SRMR = 0.04. The standardized factor loadings were all significant and generally strong (λ = 0.69–0.96; see Table 4), confirming that the constructs were well represented by their items. Scale reliabilities (Cronbach's α) were acceptable to high (see Table 1). Correlations among the latent factors ranged from 0.53 to 0.91, reflecting substantial conceptual relatedness among the organizational antecedents. Nevertheless, the separate factor structure, significant factor loadings, and the Fornell–Larcker results provided evidence that the constructs were empirically distinguishable despite this overlap. The corresponding Fornell–Larcker results are reported in Supplementary Table 10. Thus, Hypothesis 4.1 was supported, although discriminant validity should be interpreted cautiously given the high correlations among some constructs.

The structural model formulated in Hypothesis 4.2 showed a very good fit to the data, χ 2 (460) = 1288.62 ***, CFI = 0.99, TLI = 0.99, RMSEA = 0.06, SRMR = 0.03. The significant paths are displayed in Figure 4 (dotted lines show allowed correlations in the model), while all direct effects are reported in Supplementary Table 11 and the indirect effects in Supplementary Table 12.

The results indicate that peer support, feedback, and transfer capacity were positively associated with motivation to transfer, whereas supervisor support and opportunity to apply showed no significant associations. Motivation to share was positively associated with supervisor support, peer support, and feedback, while opportunity to apply and transfer capacity were not significant. Therefore, Hypothesis 4.2 was partly supported.

The aim of this study was to integrate the concepts of training transfer and knowledge sharing into a joint framework that emphasizes their complementary roles in sustaining the impact of training initiatives. In doing so, it identifies and systematizes their antecedents across individual, training, and organizational clusters and introduces an integrated approach for assessing how formal learning can stimulate subsequent informal knowledge sharing. Drawing on two complementary studies, we developed and tested a measurement framework that reliably captured these antecedents. The strong model fit indices and satisfactory factor loadings across all clusters support the psychometric adequacy of the scales and suggest their potential utility for future research and training evaluation purposes.

Our findings underscore the central role of motivation in driving both training transfer and knowledge sharing. Across the individual, training, and organizational cluster models, motivation was positively associated with volition and also showed direct associations with the respective post-training behaviors. This pattern is consistent with the proposed motivation–volition–behavior process and supports the role of motivation as an important process variable, in line with previous work emphasizing its relevance for learning transfer and performance outcomes (Holton et al., 2000; Gegenfurtner, 2009). Moreover, the results are consistent with the assumption that volition is relevant for explaining post-training behavior, supporting earlier research that conceptualizes volition as an important factor in the implementation of learning in practice (Seiberling and Kauffeld, 2017). At the same time, because motivation, volition, and behavior were assessed in the same measurement wave, these paths should be interpreted as model-implied associations rather than causal effects. The observed bivariate correlations between both constructs in Table 1 were only moderate, whereas the structural paths in the SEMs were stronger. This difference should not be interpreted as resulting from correction for measurement error alone. Rather, Table 1 reports observed zero-order scale correlations, whereas the SEM coefficients represent standardized latent regression paths estimated within a specified structural model. These coefficients are therefore conditional on the measurement model, the ordinal treatment of the indicators, the inclusion of antecedents as predictors of motivation, the direct effects of motivation and volition on behavioral outcomes, and the residual correlations allowed between the transfer and sharing process streams. Thus, the SEM paths capture model-implied process relations rather than simple bivariate associations. Accordingly, the high motivation–volition paths should be interpreted as evidence for the assumed process logic within the estimated SEMs, while the more moderate observed correlations indicate that the zero-order associations between the manifest scale scores are less pronounced. Across the three models, the motivation–volition paths were consistently strong and significant, which is consistent with the assumed process logic linking motivation and volitional regulation in both training transfer and knowledge sharing. The significant paths from volition to the respective behavioral outcomes further support the relevance of volitional regulation for explaining post-training behavior. At the same time, motivation also showed direct associations with both outcomes, suggesting that post-training behavior is related to both motivational and volitional mechanisms. The fact that for every cluster some antecedents were significant further indicates that training transfer and knowledge sharing cannot be explained by individual dispositions alone but must be understood as multicluster phenomena. Across the three models, the motivation–volition paths were consistently strong and significant, supporting the assumed process logic that motivation provides the basis for volitional regulation in both training transfer and knowledge sharing. The significant paths from volition to the respective behavioral outcomes further underscore the relevance of volitional regulation for translating intentions into post-training behavior. At the same time, motivation also showed direct effects on both outcomes, suggesting that post-training behavior is shaped by both motivational and volitional mechanisms.

Support for the antecedent hypotheses was more differentiated. Hypotheses 2.2, 3.2, and 4.2 were only partly supported, as not all antecedents significantly predicted both motivation to transfer and motivation to share. Instead, the findings point to both shared and process-specific antecedents. For example, growth orientation, training reflection, goal clarity, peer support, and feedback were relevant for both motivational processes, whereas other antecedents showed more specific associations with either motivation to transfer or motivation to share. The unexpected negative association between transfer self-efficacy and motivation to share should be interpreted cautiously and is discussed in more detail below. Overall, these findings suggest that training transfer and knowledge sharing are shaped by overlapping but not identical motivational conditions.

The model of individual-cluster antecedents shows that deficit orientation and growth orientation both played a role for motivation to transfer, but only growth orientation was significant for motivation to share. This suggests that while trainees may be motivated to apply knowledge either to overcome deficits or to pursue growth, only a growth-oriented framing encourages them to share knowledge with others. The two forms of self-efficacy showed divergent effects. Transfer self-efficacy was not significantly related to motivation to transfer but showed a negative association with motivation to share. One possible explanation is that learning and training transfer are often conceptualized and experienced as primarily individual endeavors, with relatively few participants adopting a team-oriented perspective. At the same time, a strong focus on individual application could reduce the perceived need to share knowledge with colleagues, which may explain the negative association with sharing motivation. By contrast, knowledge sharing self-efficacy did not predict motivation to transfer but was a strong positive predictor of motivation to share, underscoring the importance of self-beliefs in one's capacity to contribute to others' learning. Training reflection emerged as a strong predictor of both forms of motivation. Trainees who actively reflect on their learning are more motivated to implement new knowledge and more inclined to share it, highlighting the central role of self-reflective processes for sustaining workplace learning. Goal clarity was positively related to both motivation to transfer and motivation to share. Clearly defined objectives thus seem to stimulate not only the application of training content but also its dissemination among colleagues. This pattern is broadly in line with prior research suggesting that promotion-focused or developmental orientations are more likely than prevention-focused orientations to stimulate proactive learning behaviors and collaborative knowledge exchange (Van Yperen et al., 2011; Cerne et al., 2014). Likewise, the positive role of training reflection is consistent with research showing that reflective processing deepens learning and supports both application and externalization of knowledge (Riesen and Whitver, 2023; Sun et al., 2022). The positive effects of goal clarity also accord with earlier findings indicating that clearly understood training objectives facilitate transfer and help create a shared basis for communicating knowledge to others (Johnson et al., 2012; Van der Hoek et al., 2018). Finally, the positive effect of knowledge-sharing self-efficacy on motivation to share is in line with previous knowledge-sharing research emphasizing the importance of self-beliefs for employees' willingness to contribute what they know (Cabrera et al., 2006). Transfer self-efficacy was not significantly related to motivation to transfer, although prior literature would have suggested a positive role. In the present study, its zero-order association with motivation to transfer was very small, which may indicate that confidence in one's ability to apply training content is not sufficient, by itself, to increase motivational readiness. Rather, transfer self-efficacy may become more relevant at later stages of the process, such as maintaining effort during implementation. By contrast, the negative association between transfer self-efficacy and motivation to share was not expected based on prior literature and should therefore be interpreted cautiously. Given the non-significant bivariate correlation in Table 1 and the substantial shared variance among the individual antecedents, this coefficient may reflect a multivariate suppression-like pattern rather than a robust negative substantive relationship. At the same time, one possible explanation is that strong confidence in one's own application of training content may, in some contexts, be accompanied by a stronger individual-performance focus rather than a collective-learning orientation.

The model of training-cluster antecedents shows that content applicability was positively related to motivation to share, but only marginally significant for motivation to transfer. This indicates that when trainees perceive the content as valid and relevant, they are more inclined to pass it on to colleagues. Transfer design was a strong positive predictor of motivation to transfer but did not significantly affect motivation to share. This finding suggests that well-structured training programs primarily support the individual's intention to implement learning outcomes in their own work, whereas knowledge sharing requires additional factors beyond design, such as the content applicability. Training materials contributed positively to both motivation to transfer and motivation to share. High-quality materials thus appear to serve as a resource for both individual application and dissemination, as they may support recall, structure, and communication of content. Finally, exchange within the training group was a strong positive predictor of both forms of motivation. Opportunities to interact with peers during training seem to stimulate not only the intention to apply knowledge but also the willingness to share it afterward. This highlights the dual value of collaborative learning settings: they provide a rehearsal space for transfer while simultaneously building the social confidence needed for sharing knowledge with colleagues. These findings also correspond well with earlier research on training characteristics. The positive role of content applicability is consistent with transfer research showing that trainees are more likely to use and communicate learning content when it is perceived as relevant to everyday work demands (Holton et al., 2000; Kim and Lee, 2006; Lin, 2007). The strong effect of transfer design on motivation to transfer aligns with prior evidence that transfer-oriented instructional design facilitates the implementation of learning outcomes in the workplace (Holton et al., 2000; Renta-Davids et al., 2014; Santana-Domínguez, 2022). Likewise, the positive role of training materials supports previous findings that structured and reusable materials serve as an important resource for both later application and post-training dissemination of knowledge (Nielsen et al., 2023; Mehner and Kauffeld, 2023). Finally, the beneficial effect of exchange within the training group is in line with work suggesting that peer interaction during training enhances both training effectiveness and the continuation of learning-related exchange beyond the formal training setting (Mehner and Kauffeld, 2023).

The model of organizational-cluster antecedents shows that supervisor support did not significantly predict motivation to transfer but had a positive effect on motivation to share. This suggests that supervisors may be less influential in stimulating employees' personal intention to apply new knowledge, but their encouragement and recognition can create a supportive climate that motivates trainees to pass on what they have learned. In our context, supervisors may thus act more as multipliers, encouraging the dissemination of training outcomes within teams rather than directly fostering individual application. This interpretation is consistent with Massenberg and Kauffeld (2015), who also found that supervisor support plays a particularly important role in the context of transversal rather than technical trainings. Peer support emerged as a consistent positive predictor of both motivation to transfer and motivation to share, underlining the importance of colleagues as role models and partners in workplace learning processes. Opportunity to apply showed no significant effect on either motivation to transfer or motivation to share, despite its frequent identification as a central antecedent in the literature (e.g., Holton et al., 2000). One possible explanation is that while opportunities to apply are necessary for actual transfer and sharing behavior, they may not directly enhance motivational states; instead, their influence may unfold at later stages of the process. Given that opportunity to apply has repeatedly been found to predict transfer (e.g., Gegenfurtner, 2009; Holton et al., 2000; Kauffeld et al., 2008), this non-finding deserves further investigation. Feedback was positively related to both motivation to transfer and motivation to share. Constructive and timely feedback appears to reinforce trainees' perception that their new knowledge is valued and relevant, thereby strengthening their motivation to implement and disseminate it. Finally, transfer capacity was a strong positive predictor of motivation to transfer and showed a marginal effect on motivation to share. This suggests that organizational structures and resources enabling transfer are particularly important for fostering individual application of training content, while they may also contribute to knowledge dissemination when conditions are favorable.

Theoretical implications

The present study offers several theoretical implications for research on training transfer and post-training knowledge sharing. Learning should not be understood as an isolated event, but rather as a process that extends beyond formal training opportunities and unfolds along employees' individual learning trajectories (Mitschelen and Kauffeld, 2025; Poell et al., 2018). Within this perspective, our study emphasizes the importance of post-training knowledge sharing, a concept that has so far been underexplored despite its potential for strengthening organizational learning and ensuring that training outcomes extend beyond individual participants (Ackermann et al., 2020). By identifying antecedents that foster knowledge sharing, our findings contribute to an area that remains largely under-researched, while also underscoring its relevance for maximizing the organizational impact of training (Mehner et al., 2025).

Through our research, we have contributed to the connection between training transfer and knowledge sharing research, opening up new perspectives for both fields. Our aim is to encourage researchers in the field of training transfer to broaden their focus beyond the mere application of knowledge and to also consider the dissemination of that knowledge. Likewise, we hope to inspire knowledge sharing researchers to examine the unique situation of sharing knowledge that has been recently acquired through training.

At the same time, the results highlight that training transfer and knowledge sharing are related but distinct post-training behaviors that can be explained, in part, by shared antecedents at the individual, training, and organizational clusters. Transfer motivation and transfer volition emerged as key mediators in these processes. Volition, in particular, not only channels the effects of motivation but also sustains behavior in the face of obstacles, which resonates with recent evidence from e-learning contexts demonstrating its decisive role for successful learning outcomes (Seiberling and Kauffeld, 2025).

The joint investigation of transfer and knowledge sharing further illustrates how formal learning can spark informal learning processes within organizations. In our case, employees who participated in formal training brought newly acquired knowledge back to the workplace and shared it with colleagues, thereby enabling others to learn informally. Although the trainings examined here were formal learning opportunities rather than self-based learning formats, the observed pattern is still compatible with theoretical arguments that individually driven learning processes can foster subsequent social-based learning (Decius et al., 2025a). It also aligns with evidence showing that individuals follow distinct yet interconnected learning paths during processes such as onboarding, where formal and informal opportunities complement each other (Mitschelen and Kauffeld, 2025; Poell et al., 2018).

Finally, our study underscores the theoretical relevance of adopting a multicluster perspective. Antecedencies from each cluster exerted meaningful effects, suggesting that transfer and knowledge sharing are embedded in a broader system of contextual influences (Kauffeld et al., 2025). Learners' efforts can only unfold their full potential when organizational conditions are aligned accordingly.

Practical implications

Simply sending employees to training does not ensure transfer or knowledge sharing. Employees must be enabled to turn new knowledge into a resource for organizational development. This requires support at the individual, training, and organizational clusters (Burke and Hutchins, 2008; Grossman and Salas, 2011; Hughes et al., 2020).

At the individual cluster, employees are active agents in the transfer process. Motivation grows when they take ownership of learning, e.g., by applying proactively and clarifying goals beforehand. Seeking input from trainers, supervisors, or peers increases engagement and anticipation. Back at work, transfer is supported by sharing experiences, reflecting on outcomes, engaging in peer learning, and seeking feedback, all of which foster self-efficacy (Mehner and Kauffeld, 2023).

Training providers can strengthen transfer by anticipating barriers and using realistic, applicable content. Alternating learning and application phases, follow-up modules, and (virtual) coaching maintain motivation. High-quality, adaptable materials deepen learning and support application. Pre-training meetings to clarify goals also enhance transfer, though trainers often underestimate the impact of training design (Wisshak et al., 2025).

Organizations shape transfer by choosing providers carefully, reviewing didactic concepts, and requesting evidence of transfer-oriented design. Feedback from participants and evaluation of materials ensure alignment with organizational needs. Providing resources like time, tools, and structural support enables application, while a constructive error culture boosts self-efficacy. Sending employees in tandems or groups further fosters exchange and prevents isolation (Burke and Hutchins, 2008; Hughes et al., 2020; Salamon et al., 2023).

In addition, systematic diagnostics can help identify barriers and catalysts for transfer and knowledge sharing at an early stage. Assessing the identified antecedents of training transfer and post-training knowledge sharing can serve as an early warning system before large-scale training initiatives, support the evaluation of existing programs, diagnose causes of transfer problems, guide targeted interventions, and raise awareness among trainers and supervisors of transfer-related challenges. Depending on the objective, antecedents can be measured before training as a readiness check, directly after training for evaluation, several weeks later as a transfer check, or more broadly to assess the overall transfer climate within an organization. Presenting results in terms of barriers and catalysts clearly shows where action is needed to improve both training design and its workplace implementation.

Taken together, these implications underscore that training transfer is not a one-time event but an ongoing process that requires shared responsibility among employees, colleagues, supervisors, organizations, and training providers. By systematically addressing both motivational and volitional aspects, and by embedding trainings in supportive organizational structures, organizations can move beyond the notion of merely “sending employees to training.” Instead, they can foster sustainable learning cultures that enable knowledge to be continuously applied, adapted, and shared. Importantly, employees and organizations alike benefit when different forms of learning are meaningfully connected along individual learning paths. In this way, training not only supports individual growth but also becomes a driver of collective learning and organizational development.

Limitations and future research

The joint investigation of training transfer and knowledge sharing provides new insights into how formal learning can lead to informal learning processes within organizations. At the same time, Mehner et al. (2025) showed that knowledge sharing may itself promote further informal learning among training participants. They also argued that knowledge sharing does not necessarily follow transfer but may occur in parallel. Our findings support this interpretation, yet the precise temporal order remains an open question. In addition, our design did not model direct paths between training transfer and knowledge sharing, which limits conclusions about their causal or sequential relationship. Future research should therefore examine whether applying knowledge first strengthens the capacity to share, whether sharing accelerates the individual application of training content, or whether both processes reinforce each other in parallel. Moreover, knowledge sharing captures only one aspect of informal learning. Future studies should therefore also consider other informal learning processes, such as reflection, feedback, and exchange with colleagues, to better understand how formal and informal learning interact over time.

Additionally our sample size, as determined by a G * power analysis, was unfortunately not sufficient to test the direct effects on volition in addition to the effects we tested. For this reason, we focused on the prediction of motivation and modeled the indirect pathways. Since our results underscored the crucial role of volition, future studies should integrate volition more systematically and investigate the direct effects of antecedents on volition.

Another limitation concerns the composition of the Study 2 sample. A substantial proportion of participants reported holding a leadership or managerial position, and the sample was relatively highly educated. This may limit the generalizability of the findings to employees without leadership responsibilities or with different educational backgrounds. Leadership or managerial roles may shape how participants perceive supervisor support, peer support, opportunities for knowledge sharing, and their own responsibility to disseminate training content. Similarly, educational background may be related to prior learning experiences, self-efficacy, and familiarity with formal training contexts. Future research should therefore replicate the findings in more balanced samples and examine whether the proposed relationships differ between managerial and non-managerial employees or across educational groups.

The present study estimated separate structural models for the individual-, training-, and organizational-level antecedents. This approach was chosen to retain the construct-specific interpretation of the antecedents within each cluster and to test whether these conceptually related constructs can be empirically distinguished and linked to the training transfer and knowledge sharing processes. At the same time, this segmented approach limits conclusions about the simultaneous and relative predictive strength of the three antecedent levels. Thus, the present findings do not show whether individual, training-related, and organizational antecedents explain unique variance when modeled jointly, nor do they capture potential cross-level dynamics among these antecedent domains. Future research could build on the present findings by testing a fully integrated model, for example using higher-order specifications or PLS-SEM, to examine how the three antecedent levels jointly predict motivation, volition, training transfer, and knowledge sharing. Such analyses should ideally be conducted with matched samples using identical item sets. In the present research, Study 1 served item development and refinement, whereas Study 2 used the reduced and translated item set. This limits the suitability of the two datasets for formal multi-group comparisons or measurement invariance testing across the German-speaking and U.S. samples.

A further limitation is that the antecedents may differ in their temporal and conceptual character. Some constructs, especially within the individual and organizational clusters, may reflect relatively stable orientations or contextual perceptions, whereas training-related antecedents as well as motivation, volition, training transfer, and knowledge sharing are more closely tied to the focal training episode. Although all measures referred to the same training context, this distinction was not modeled statistically and should be addressed more explicitly in future research.

Another limitation concerns the reliance on self-report data. All main variables, including antecedents, motivation, volition, training transfer, and knowledge sharing, were assessed from the participants' perspective. This approach was appropriate because the study focused on trainees' perceptions of their training context and post-training behavior. However, it may also introduce common-method bias and may be affected by social desirability or individual response tendencies. Although the two-wave design temporally separated the assessment of antecedents from motivation, volition, and behavioral outcomes, motivation, volition, training transfer, and knowledge sharing were assessed within the same later wave. Therefore, associations among these variables should be interpreted with caution. Future research should complement self-report data with additional sources, such as supervisor or peer ratings, behavioral indicators, or organizational records.

One strength of our study is its two-wave, time-lagged design, in which antecedents were assessed 1 month prior to outcomes. This design allowed us to test predictive relationships. At the same time, a key limitation is that participants reported on trainings that had taken place several months earlier on average. While this is not inconsistent with transfer research, which conceptualizes transfer as behavior maintained over time (Wisshak et al., 2025), the retrospective design may have introduced recall bias and reduced temporal precision. Moreover, the chosen 1-month lag between antecedents and outcomes may not fully capture the longer-term dynamics of transfer and knowledge sharing. Future research should therefore combine assessments immediately after training with multiple follow-up measurements in order to generate more precise insights into the trajectory of transfer and knowledge sharing over time. Relatedly, Massenberg et al. (2017) and Gegenfurtner (2009) showed that different antecedents may exert their influence before vs. after training. Extending such analyses to additional antecedents and multiple time points would substantially refine theoretical understanding. In this regard, it would also be valuable to compare different types of training, such as technical vs. transversal programs, since there is some evidence that antecedents differ between them. Moreover, future work could build on recent conceptual advances in self-regulated learning (Decius et al., 2025b) to deepen understanding of how learners actively manage transfer and knowledge sharing processes.

Moreover, future research should examine cultural differences in how antecedents operate. Labor market conditions in the U.S. and Germany are increasingly diverging, for instance in the duration of job tenure. National statistics show that the median job tenure in the U.S. was just 3.9 years in January 2024, the lowest level since 2002 (Bureau of Labor Statistics, 2024). In contrast, in Germany 41.7% of employees had been with their current employer for at least ten years in 2023, and another 20% for 5 years −10 years (Statistisches Bundesamt [Destatis], 2023). Such significant differences in job stability are likely to influence training transfer and knowledge sharing behaviors, as frequent job changes in the U.S. may limit long-term effects or alter the relative importance of individual antecedents. Frequent supervisor changes in more volatile labor markets, for example, may reduce the impact of supervisor support. While many antecedents have shown similar effects across Europe and the U.S. in the past, it remains an open question whether these patterns still hold under the current, increasingly divergent labor market conditions.

In sum, this study contributes to the theoretical understanding of training transfer and post-training knowledge sharing by providing systematically developed and empirically tested scales, confirming the role of motivation and volition, and underscoring the importance of examining antecedents across multiple clusters. By including both training transfer and knowledge sharing in a joint model, and by demonstrating that both are explained by shared mechanisms, our study advances theory and measurement in this underexplored field. By integrating formal and informal learning perspectives into a unified model, it takes an important step toward a more comprehensive understanding of sustainable workplace learning.