Development and validation of the learning culture scale for small- and medium-sized enterprises

Technological developments, shifting markets, changing demographics and unexpected disruptive events continuously challenge organizations to innovate. An important prerequisite for innovation is continuous employee learning (Achdiat, Mulyani, Azis, & Sukmadilaga, 2023; Sung & Choi, 2014). As employees increase their skills and knowledge through learning, organizational innovation and performance increase (Jacobs & Washington, 2003; Kareem & Hussein, 2019), as well as employability of employees (Yang, Watkins, & Marsick, 2004) and work engagement (Malik & Garg, 2020). While the importance of employee learning is evident, creating an environment in which continuous employee learning truly takes place is a significant challenge for organizations, especially for SMEs (Short & Gray, 2017). Several factors contribute to why learning and development are characterized as less sophisticated and insufficient compared with larger firms. First, managing daily routine operations on the one hand, while anticipating market changes on the other hand is particularly challenging for SMEs (March, 1991) leaving little room for employee learning. Second, employees within these SMEs, many of which are vocationally educated, share a negative perception of learning (Decius, Schaper, & Seifert, 2021) which potentially limits learning within these organizations. Third, SMEs seem to lack the expertise and resources to design a working environment in which employee learning is facilitated (Coetzer et al., 2017; Harney, Gilman, Mayson, & Raby, 2022). Furthermore, quantifying the return on investment for employee learning is challenging, making it especially daunting for SMEs with limited resources (Altinay, Madanoglu, De Vita, Arasli, & Ekinci, 2016).

Organizations can set out to increase employee learning by creating positive conditions within the organization that facilitate the learning of individuals through an atmosphere that enables and creates incentives for collaboration, reflection and inquiry (Örtenblad, 2002). Such a learning culture has been shown to be beneficial for both the employee and the organization (see for example Egan, Yang, & Bartlett, 2004; Liao, Fei, & Liu, 2008; Sung & Choi, 2014). However, the conceptualization of learning culture, its measurement instruments and the potential beneficial outcomes have been studied particularly in larger organizations (e.g. Marsick & Watkins, 2003; Škerlavaj et al., 2007; Yang et al., 2004, etc.). The extent to which the underlying principles and outcomes of these studies are applicable to SMEs is unknown. The specific characteristics of SMEs render it plausible that learning culture takes on different shapes and forms compared to larger organizations (Coetzer et al., 2017). Consequently, alternative methods to shape a learning culture may be needed. The significant role of SMEs in the general economy and the employment opportunities they provide, the importance of employee learning within SMEs and the specific SME characteristics that limit the applicability of prior studies, renders the exploration of how a learning culture can be captured in SMEs a worthwhile endeavor.

The main goal of the current study is therefore to present the conceptualization and validation of a scale that can assess the learning culture within SMEs. Here, a learning culture scale for SMEs was conceptualized using the literature on learning organizations (Marsick, 2013; Marsick & Watkins, 2003; Yang et al., 2004) and learning and development within SMEs (Coetzer et al., 2017, Coetzer, Wallo, & Kock, 2019) and validated using survey data from 59 SMEs in two rounds of data collection.

Organizational learning culture refers to the shared values, attitudes, beliefs and practices within an organization that promote and support continuous learning and improvement (Van Breda-Verduijn & Heijboer, 2016). In other words, it is the shared collective mindset and behavior of an organization that supports the acquisition of information, the distribution and sharing of learning and reinforces and supports continuous learning and its application to organizational improvement (Bates & Khasawneh, 2005). Organizational learning is associated with many positive outcomes. For one, it has been shown to affect employee job satisfaction in a positive manner (Lund, 2003). In addition, Škerlavaj et al. (2007) found that organizational learning culture affects business financial performance, mediated by employee performance. Furthermore, Ellinger et al. (2002) showed a positive correlation between organizational learning culture and the financial performance of firms, explaining 10% of its variance. More recently, a comprehensive literature review shows that organizational learning culture is an important enabler for innovation (Achdiat et al., 2023). As a body of literature increasingly illustrates the positive outcomes of a learning culture, interest in capturing the learning culture of an organization has grown. In the present study, we contribute to the body of knowledge by presenting the conceptualization and validation of the Learning Culture Scale (LCS) that focuses on capturing the learning culture within an SME.

Conceptualizing the LCS was done using two theoretical perspectives: the analytical framework of the learning organization developed by Marsick & Watkins (1990, 2003) and more recent insights on learning within SMEs (Coetzer et al., 2017, 2019; Csillag, Csizmadia, Hidegh, & Szászvári, 2019). The analytic framework of the learning organization (Marsick & Watkins, 1990, 2003) has several important features that suit the conceptualization and validation of a learning culture scale. First, the framework is one of few that encompasses most learning concepts relevant in literature (Örtenblad, 2002) and captures those concepts in underlying dimensions. This allows for the LCS presented here to focus on the most salient dimensions that cover the underlying concepts without the need to untangle and specify all those individual concepts. Second, the framework integrates the dimensions by specifying how the dimensions are interrelated. This allows for a holistic perspective on learning culture. Thirdly, the framework enables a multilevel perspective on learning culture, as it encompasses four levels of a learning organization:

1. the individual level, comprised of two dimensions of organizational learning: continuous learning and dialogue and inquiry;

2. the team or group level, which is reflected by team learning and collaboration;

3. the organizational level consisting of two dimensions of organizational learning: embedded systems and empowerment; and

4. the global level, which consists of two dimensions of organizational learning: systems connection and strategic leadership.

Hence, the framework underlines that learning takes place at the individual, team and organizational level, and therefore ought to be facilitated by the organization on all these levels.

Recent insights showed that learning and development in SMEs differ from large organizations (Coetzer et al., 2017, 2019; Csillag et al., 2019). The traditional approach to promoting learning by providing employees access to employer-funded training (i.e. formal learning) is often criticized as an “ineffective” approach and an “inappropriate” way of conceptualizing learning in SMEs (Anderson & Boocock, 2002; Billett, Hernon-Tinning, & Ehrich, 2003). Rather than relying on traditional ways of learning (e.g. attending training courses), SMEs should develop an approach congruent with the distinctive characteristics of the small firm (Kelliher & Henderson, 2006). Coetzer et al. (2017) make a distinction between task, relational and organizational features of a learning organization in the context of SMEs. Initiatives such as harnessing the learning potential in daily work practices, team-based learning practices (Ahlgren & Engel, 2011; Birdthistle, 2008; Coetzer et al., 2017; McPherson & Wang, 2014), local learning support (Billett et al., 2003), incentives and support for learning (Coetzer, 2006; Ehrich & Billett, 2004) and enhancing the learning potential of the work system by increasing the levels of autonomy and skills variety in jobs (Coetzer, 2006) are examples of task, relational and organizational features of learning within an SME. Furthermore, the roles of business owners, managers and other potential learning facilitators (e.g. peers, advisers, experts, family members) are important in creating conditions that foster learning within SMEs (Billett et al., 2003; Coetzer, 2006; Dalley & Hamilton, 2000; Kelliher & Henderson, 2006).

The present study integrates the existing conceptualization of the learning organization (Marsick, 2013; Marsick & Watkins, 1990, 2003) with specific elements that foster learning in SMEs (Coetzer et al., 2017, 2019). What these theoretical perspectives have in common is that learning culture is regarded as a multidimensional construct that encompasses both elements of behavioral and structural factors that foster learning in SMEs. Furthermore, both frameworks define the elements that foster learning on different levels (individual, group/team and organization; see also Alves & Thiebaut, 2024). By integrating them we identified six dimensions of learning culture relevant to SMEs.

Learning-oriented tasks refer to the extent to which employees perceive their organizations to allocate tasks that enable continuous learning. Many tasks in SMEs tend to be characterized by high task- or skill variety, a wide scope for action and high task significance. Therefore, learning is enhanced by embedding learning in everyday goal-directed work activities (Coetzer et al., 2017). Such activities require employees to step outside their comfort zone while expanding their knowledge and skills to find solutions to the challenges at hand and learn (Nurmi & Hinds, 2016). To enhance learning from these tasks, it is important that employees have sufficient autonomy. Task autonomy ensures that employees have the opportunity to come up with a solution for the challenges themselves (Morgeson & Humphrey, 2006). In doing so, they develop their problem-solving skills. Autonomy also allows employees to regulate and manage the pressure of complex work and set the right priorities through which they truly learn (Van Ruysseveldt & van Dijke, 2012). The variety of activities also contributes to employee learning (Parker, 2014). By undertaking various tasks, employees are exposed to different skill and knowledge demands, which may require them to learn new things.

Reflexivity and experiential learning refers to the extent to which employees perceive they are encouraged to experiment and reflect on their work and engage in (open dialogue about) learning from mistakes. This concept is closely related to elements of psychological safety (Edmondson, 1999) which state that learning behavior can be promoted in organizations by reducing excessive concerns about others’ reactions to initiatives that may lead to embarrassment or threat. As such threats are often experienced during learning initiatives, employees must be encouraged to try out new working methods, unlearn routines and experiment without the risk of being held accountable for it (Costigan, Iiter, & Berman, 1998). Stimulating such behaviors in organizations results in employees who are more likely to take calculated risks, are more likely to try things out and develop themselves (Borgatti & Cross, 2003). Another aspect of this dimension is giving and receiving constructive feedback (Van Woerkom, 2004). Especially if there is room for experimentation, feedback loops are detrimental to increasing learning potential (Tannenbaum, Beard, McNall, & Salas, 2009). Feedback moments can be built into standard working methods or supported with training. From previous studies we that psychological safety and reflexivity enhance knowledge sharing among employees (Curado & Vieira, 2019) and innovative capabilities and performance in SMEs (Andersson, Moen, & Brett, 2020).

Collaborative learning refers to the extent to which employees perceive their organizations to facilitate learning from each other. Learning is a social process that is often stimulated by working with other people (Kayes, Kayes, & Kolb, 2005). This requires structural opportunities to share experiences and ideas with each other informally and in an accessible manner (Levine & Moreland, 1999; Rupert & Jehn, 2008). In SMEs, there is a greater likelihood of close, informal and accessible interactions, which creates a conducive environment for learning (Coetzer et al., 2017), providing great potential for vis-à-vis learning. An example of such an informal way of learning is mentoring, in which an existing employee can guide newcomers or less-experienced employees in a task to develop professional skills, attitudes and competencies (Hamburg, Brien, & Engert, 2014; Kram, 1985). In transferring tacit knowledge and retaining expertise within the organization, but also by passing on corporate vision and values, mentoring can also help the mentee become a recognized and accepted member of the community. Another important element of collaborative learning – especially in the SME context – is knowledge sharing across teams. As tasks and roles are less likely to be formalized within SMEs (Coetzer et al., 2017) and task and role interdependency is higher (De Clercq, Dimov, & Thongpapanl, 2015), the need to exchange knowledge between team or organizational units is more important to reach organizational goals. Hence, the potential for learning through sharing knowledge across teams is high within SMEs.

Learning-oriented leadership refers to how employees perceive the leadership within the organization as conducive to learning. On the one hand, this refers to having a clear vision of how strategic challenges that the organization faces can be aligned with learning and development goals, and the ability to identify and communicate what is needed on the employee level (Marsick & Watkins, 2003). On the other hand, learning-oriented leadership refers to behaviors that encompass facilitating and supporting learning processes among individuals and teams/departments. Coetzer et al. (2017) label the low levels of formality between leaders and employees within SMEs as detrimental to increasing learning potential among employees. In addition, low formality interactions increase the impact on the role modeling function that managers have in fostering learning behavior of employees (Hillberg Jarl, 2024; Lyons & Bandura, 2020).

Organizational learning structure refers to the extent to which employees perceive the organizations to provide means (time, resources, tools) to facilitate learning, and whether the organization provides structures in which learning can be enhanced. It is essential to organize the organization in such a way that allows learning experiences to be shared, and also formally integrated into the structure of the organization (Marsick & Watkins, 2003). Firstly, this refers to how workplaces and spaces are designed to increase the learning potential of individual employees (Parker, 2014). Secondly, it encompasses how the opportunities for knowledge sharing are formally organized within the organization. For example, by organizing daily stand-ups or “huddles” for communicating and solving specific problems relevant to a whole team or department (Ankomah, Ayarkwa, & Agyekum, 2017). This seems especially challenging in SMEs as these organizations often are less hierarchical and more fluidly structured (Coetzer et al., 2017; Goh, 1998). Finally, organizational learning structure concerns the provision of information about the possibilities for learning and development within the organization. It is essential that employees are aware of the possibilities and meaning of learning resources (cf. Dymock & McCarthy, 2006).

External network learning refers to the extent to which the employees perceive that the organization facilitates learning from outside sources. When employees are in contact with others who are not part of the organization, it offers new arrays for learning (Humphrey, Nahrgang, & Morgeson, 2007). Interaction with people outside the organization with which they exchange and share implicit and explicit knowledge increases task and role complexity (Berman, Down, & Hill, 2002) which enables learning. For example, sharing requirements, challenges and questions with suppliers or customers results in better insight into trends and developments in the sector and how this relates to employees’ daily work (Alerasoul, Afeltra, Hakala, Minelli, & Strozzi, 2022). In addition, it is also important that employees proactively scan the environment, join relevant networks and retrieve and use information smartly (Gerken, 2016). In this regard, Oh & Kim (2022) showed that among SMEs using knowledge of external organizations leads to the accumulation of knowledge at the individual group and organizational level, which directly influences organizational innovation.

In the present study, we present the development and validation of the Learning Culture Scan (LCS) based on these six dimensions. The development and validation process for the LCS was conducted in two separate consecutive studies. In Study 1, items for the LCS were developed based on the literature and additional semi-structured interviews among representatives of SMEs. This provided insights into the specific context of SMEs and helped to tailor the items to the practical realities of these organizations. The item set generated in these two steps was subsequently exploratory tested for its underlying factor structure in a sample of employees from SMEs and non-fitting items were omitted. Next, in Study 2, the reduced item set was further tested for its internal factor structure in a larger sample of employees from a larger group of SMEs. Additionally, Study 2 examined its relationship with relevant outcomes, such as the effect on the amount of time spent on formal courses or training in the last two years. The following sections outline the approach, findings and implications of Study 1 and Study 2.

The purpose of Study 1 was twofold: to develop a self-reported scale for assessing the learning culture within SMEs, and to conduct an initial exploration of the statistical validity of the questionnaire at hand. We adhered to the guidelines for scale development outlined by Hinkin (1995), who divides this process into three phases:

1. item generation;

2. scale development; and

3. scale evaluation.

The Learning Culture Scan (LCS) aims to assess the learning culture in SMEs across the six dimensions previously described, using a standardized approach to item formulation. The developed items were generated based on insights from:

• learning organization and learning in SMEs (Marsick & Watkins, 2003; Yang et al., 2004) and learning and development in SMEs (e.g. Coetzer et al., 2017, 2019); and

• through 36 semi-structured interviews within SMEs [board members (11), HR managers (9), line managers (7) and employees (9) from ten Dutch SMEs with a size ranging from 10–100 employees].

The goal of the interviews was twofold. First, we checked whether the representatives from the SMEs recognized the six dimensions as important to their organization to facilitate learning behavior among employees. Second, we asked them to provide practical examples of prototypical activities for each of the dimensions. For example, with regard to reflexivity and experiential learning, several participants described informal work meetings or “day starts” in which employees shared mistakes to learn from. This information was used to gather input for generating additional items to ensure that the content of the measurement instrument not only fits the theoretical conceptualization of the dimensions but also relates to the context of the daily work of employees in SMEs.

After the first item generation, based on the two previously described steps, all items were subjected to reflective discussions between the researchers. This resulted in a list of 59 items to assess the learning culture in SMEs, with responses plotted on a five-point Likert scale (value 1 = strongly disagree, 2 = disagree, 3 = neutral, 4 = agree, 5 = completely agree). Table 1 provides a detailed overview of the questionnaire structure and underlying sources.

To evaluate the scale, we collected data from 388 employees of the 12 Dutch SMEs (size from 10 to 100 employees). The companies are active in different sectors: industry (n = 5), information and communication (n = 5), trade and transport (n = 1) and business services (n = 1). For each participating organization an introductory meeting was organized to explain the aim of the study. The LCS was distributed to participants via an anonymous Web link or completed on a laptop in the workplace. Completing the questions took 15–20 min. Informed consent was included as part of the questionnaire. Within the SMEs, participants showed an average response rate of 55% over the total sample size, with responses ranging from 39% to 100%. A total of 83% achieved a response rate of more than 50%.

After data collection, quantitative analyses were applied to evaluate the validity of the scale constructs. Reliability and validity were evaluated using IBM SPSS version 27 and JASP 0.16. First, we performed the descriptive analysis of the dimensions of the LCS (means, standard deviation, skewness and the shape of the response distribution). Secondly, as a measure of reliability, Cronbach’s alpha coefficient was calculated based on the six dimensions of the learning culture to determine whether the items are sufficiently interrelated. Finally, a confirmatory factor analysis was conducted to assess the extent to which the theoretical constructs (the six dimensions of a learning culture) were empirically supported. Four indicators were used to determine the fit of the model. Common cutoffs for fit indices were applied to evaluate model fit (Hu & Bentler, 1999). The Comparative Fit Index (CFI) and the Tucker–Lewis Index (TLI) were considered measures of model fit and reliability, with CFI and TLI values greater than 0.90 considered acceptable to good. The Root mean square error of Approximation (RMSEA) was examined as an approximation error index, with an RMSEA value of <0.06 indicating a good fit (Hu & Bentler, 1999). Browne & Cudeck (1993) suggest a range between 0.05 and 0.08 for a reasonable fit. Finally, the Standardized Root Mean Square Residual (SRMR) was included, which assesses the discrepancy between the observed correlations and the correlations predicted by the model. SRMR is considered acceptable if it is < 0.08 (Hu & Bentler, 1999).

Table 2 shows the descriptive statistics for the six dimensions of learning culture. It can be observed that the average scores range from 3.08 (external network learning) to 3.46 (reflexivity and experiential learning). The standard deviation (SD) varies across the dimension from 0.40–0.75. Skewness values range between −0.293 and −0.837, indicating a slight to moderate negative skew in the distribution of data for the different elements of learning culture. Kurtosis ranged from −0.107 to 0.995, indicating a symmetrical distribution of responses. To assess the reliability, Cronbach’s alpha coefficient was calculated for the 0-model (baseline model) consisting of 59 items. The internal consistency reliability of the six learning culture dimensions appeared to be sufficient – except for the dimension learning-oriented tasks – based on the guidelines (Nunnally & Bernstein, 1994; Vaske, Beaman, & Sponarski, 2017), who consider an α ≥ 0.60–0.65 acceptable.

Subsequently, a confirmatory factor analysis was conducted on the total data set of Study 1 (n = 388). The results of the chi-square test on the basic six-dimensional model showed that the factor model (χ² = 3890.505, df = 1637, p < 0.001), fits the observed data significantly better than the baseline model. The results of the fit indices showed that the basic model with 59 items needs adjustments to achieve a better fit. The fit indices suggest mixed model fit: while the CFI and TLI fall below the threshold of 0.90 (CFI = 0.792; TLI = 0.783), the RMSEA indicates reasonable fit, ranging from 0.057 to 0.062 and the SRMR is acceptable at 0.060 (Browne & Cudeck, 1993; Hu & Bentler, 1999).

We examined possible improvements in the model fit step by step through iterative adjustments. To determine which items should be removed from the basic model (59 items), we first examined the factor loadings. Items with an estimated factor loading below 0.3 were excluded, because they did not make a substantial contribution to the model (Field, 2013). Subsequently, items with factor loadings between 0.3 and 0.5 underwent thorough analysis by the researchers, using both statistical criteria and substantive interpretation (Hair, Black, Babin, & Anderson, 2019). This included examining both the factor loadings and the wording of the items. If the wording of an item was very similar to that of an item with a higher factor loading, we determined which of the two items best matched the dimension in question. In addition to evaluating the factor loadings and their p-values, we also considered possible cross-loadings, checking whether variables loaded significantly on multiple factors by carefully inspecting the parameter estimates from the confirmatory factor analysis. Furthermore, reliability measures were used to assess whether removing items resulted in a higher Cronbach’s alpha coefficient, as a measure of the internal consistency of the measuring instrument. This process led to a reduction in the number of items in the basic model from 59 to 47 (see  Appendix for the items for each learning culture scale dimension).

Table 3 shows the model fit indices of both the basic 59-item learning culture model and the improved 47-item model. In the improved model, the CFI and TLI values increased to 0.823 and 0.812. The RMSEA and SRMR values have maintained their good fit.

Table 4 presents the descriptive statistics and reliabilities of the final model and shows that the adjusted model with 47 items contains similar patterns and variations compared to the values presented in Table 2. However, the reliabilities of the subscale “learning-oriented tasks” have significantly increased to 0.661 and show acceptable internal consistency (Nunnally & Bernstein, 1994; Vaske et al., 2017).

The purpose of Study 2 was to further validate the revised questionnaire with a larger sample of SMEs, assess its relationship with several demographic variables and explore the predictive validity of our instrument. This contributes to a broader understanding of the generalizability and applicability of the revised questionnaire for assessing perceived learning culture in different contexts. For Study 2, we followed the same procedure as in Study 1. To validate the robustness and hierarchical structure of the model, analyses were conducted on a larger sample. As in Study 1, Study 2 included CFI, TLI, RMSEA and SRMR as indicators for comparative fit. In addition, we measured gender, age and tenure of the participants. Finally, the predictive validity was assessed by looking at the effect of the six dimensions of learning culture on the amount of time spent following a course or training in the past year.

A total of 1,437 employees from 47 different Dutch SMEs (size from 10 to 100 employees) filled out the LCS. The companies are active in different sectors: industry (n = 31), construction (n = 11), trade and transport (n = 3) and business services (n = 2). The average response rate among participants in Study 2 was 62% of the total sample size, ranging from 26% to 100% of the total sample size. Of the organizations surveyed in Study 2, 81% achieved a response rate of more than 50%.

The mean scores from Study 2 are highly consistent with the results from the improved model from Study 1 (see Table 5). In Study 1, two items were removed from the “learning-oriented tasks” dimension. This led to an improvement in the alpha coefficient in that study (α = 0.661). This improvement was sustained in Study 2 (α = 0.662), which exceeds the generally accepted threshold of α > 0.60–0.65 (Nunnally & Bernstein, 1994; Vaske et al., 2017).

Subsequently, we calculated whether the structure with the six dimensions of learning culture, measured in the basic model via 47 items, could be confirmed with the collected data. The confirmatory factor analysis showed - in line with Study 1 – an acceptable fit (χ² = 6024.33, df = 1019; χ²/df = 5.91). Both the CFI and TLI were 0.828 and 0.817, respectively, below the threshold of 0.90 and slightly lower compared to Study 1. The RMSEA in Study 2 was 0.057–0.060, indicating a smaller deviation between observed and modeled covariances. Finally, the SRMR was 0.051, indicating a slightly better fit compared to Study 1. Figure 1 displays the standardized factor loadings of the final model from Study 2.

No gender effects were found with regard to the six dimensions of the LCS. We found small age and tenure effects on several dimensions of the LCS. Age was negatively correlated with learning-oriented tasks (r = −0.11, p < 0.001), reflexivity and experiential learning (r = −0.11, p < 0.001) and external network learning (r = −0.06, p < 0.05). The same pattern was found with regard to tenure. Tenure was negatively correlated with learning-oriented tasks (r = −0.09, p < 0.001), reflexivity and experiential learning (r = −0.13, p < 0.001) and external network learning (r = −0.08, p < 0.01). Older employees that worked longer in their organization perceive their tasks as less learning oriented, experience less room for experimentation and constructively reflect on one’s mistakes on the work floor and regard their organization as less compared to younger employees.

By conducting a linear regression analysis on the data from Study 2 (n = 1437), the relationship between the six dimensions of learning culture and the effect on the amount of time spent on formal courses or training in the past two years. The analysis of variance (ANOVA) showed a significant relationship between the time spent on courses or training and the six dimensions of the learning culture [F (6, 1428) = 9.84, p < 0.001]. The regression coefficients then indicate how the six dimensions of a learning culture are related to the time spent on courses or training. The results show that the “learning-oriented leadership” dimension has a significant positive influence (B = 0.68, p < 0.001). In other words, the more employees perceive their management as learning-oriented, the more time they spend on formal training. In contrast, the “collaborative learning” dimension has a significant negative influence on the time spent on courses or training (B = −0.41, p = 0.004). Thus, the more employees perceive that in their organization people learn from each other, the less they spend time on formal forms of learning. In contrast, the other dimensions showed no significant relationship with the time spent on (formal) courses or training.

Technological developments, shifting markets, changing demographics and unexpected disruptive events continuously challenge organizations to innovate. An important prerequisite for innovation is continuous employee learning (Achdiat et al., 2023; Sung & Choi, 2014). However, creating an environment in which continuous employee learning truly takes place presents a significant challenge for organizations, especially in SMEs. Organizations can set out to increase employee learning by creating positive conditions within the organization that facilitate the learning of individuals through an atmosphere that enables and creates incentives for collaboration, reflection and inquiry (Örtenblad, 2002). However, the dimensions of such a learning culture and the potential beneficial outcomes have been studied particularly in large organizations The present study therefore aimed to develop a scale that measures learning culture in SMEs. In two studies the 47-item scale was validated among 1.825 employees in 59 SMEs.

Confirmatory factor analyses and analysis of intercorrelations provide moderate to strong support for the internal structure and validity of the LCS. More specifically, in Study 1 we found that our 59-item measure of learning culture taps into six factors and showed good to moderate internal reliability. However, a trimmed 47-item solution increased model fit and internal reliabilities substantially. Confirmatory factor analysis of the trimmed 47-item scale in a larger sample Study 2 replicated the distinction between six factors of learning culture, providing further support for the internal validity of the LCS.

An examination of the LCS’s associations with other variables showed that learning culture is experienced differently by different subgroups. Both younger and less tenured workers experienced their learning culture stronger in terms of the presence of learning-oriented tasks, reflexivity and experiential learning and external network learning. These results are in line with previous findings in the literature on learning culture related to age. For example, Coetzer (2007) found that younger employees viewed both work environment conditions and their immediate supervisor’s level of proximate support for learning more favorably than relatively older employees did. In addition, younger employees generally reported greater satisfaction with their learning experiences in daily tasks than relatively older respondents did. The LCS also showed predictive value on actual learning behavior in terms of the amount employees engaged in formal courses and training. We found this relationship particularly strong for those employees who perceive their managers as learning-oriented in their leadership style. The collaborative learning dimension, which emphasizes direct learning from colleagues and knowledge sharing, is, however, negatively related to formal training. This may seem counterintuitive, but it may be explained by considering the fact that in collaborative learning, direct learning from others and knowledge sharing are important features, while leadership-focused learning can be seen more as an incentive or stimulus for learning aimed at the development of both individuals and teams. It is conceivable that employees who perceive their workplace as highly focused on collaborative learning may have less need for formal courses and training, as their learning needs are already being fulfilled in other ways within the team. Additionally, in leadership-focused learning, leaders are regarded as facilitators of the learning process. This raises the question of whether SMEs always designate individuals to fulfill this role and whether this is considered a primary or secondary responsibility by them.

Learning culture is an important and well-researched theme in HRD literature. The unique theoretical contribution of this study is that we integrate insights from more general learning organization literature with insights from learning and development in SMEs. The traditional approach to promoting learning by providing employees access to employer-funded training (i.e. formal learning) is often criticized as a too-limited way of conceptualizing learning in SMEs (Anderson & Boocock, 2002; Billett et al., 2003; Coetzer et al., 2019). In the present study, we corroborate that the distinctive characteristics of the small firm should reflect the task, relational and organizational features of a learning organization (cf. Coetzer et al., 2017). These features, which include initiatives such as harnessing the learning potential in daily work practices, team-based learning practices, local learning support incentives and support for learning and enhancing the learning potential of the work system by increasing the levels of autonomy and skills variety in jobs, as well as the role of business owners, managers and other potential learning facilitators, are all captured in the LCS.

Furthermore, this study is, to our knowledge, a first attempt to develop an instrument measuring the construct of the learning culture within SMEs with appropriate psychometric procedures. It contributes to the literature by developing a conceptual definition of learning culture in SMEs, measuring it, providing evidence of its construct validity and demonstrating its relationship with time spent on formal training. The latter finding corroborates previous studies showing that learning climate is related to increased workplace learning in organizations (see Nikolova, Van Ruysseveldt, Van Dam, & De Witte, 2016). However, future research could expand upon examining relationships between the learning culture dimensions and other outcomes. Potential studies could for example be targeted at how these dimensions of learning culture relate to productive and affective outcomes on the individual level, such as job performance (Nam & Park, 2019) or organizational commitment (Joo & Shim, 2010) but also financial performance or growth on the firm level (see Ellinger et al., 2002). Furthermore, the LCS could be used as a longitudinal instrument for measuring learning culture changes over time or used it as a pre and post-measure to examine to what extent certain learning interventions yield in terms of structural changes in learning behavior on the work floor.

This study has a few limitations. First, even though we used data from a heterogeneous sample of SMEs the data collected is skewed towards SMEs in the technical industry sector. Hence, future studies might benefit from testing our instrument using more independently collected samples. This would allow for the LCS to be cross-validated across samples of sectors whose samples were not included in the current study. In addition, our conceptualization and item generation are fitted in the context of SMEs. However, the question remains whether our measure is specifically only applicable to SMEs or also (partly) generalizable to other contexts. Our study would have benefitted from a comparison of the factor structure in larger organizations.

Furthermore, our study relied on cross-sectional data, which does not allow us to establish the directionality of the relationships of the learning culture dimensions with time spent on formal training. Even though organizational culture is generally considered to be a trigger of employee outcomes (Schneider, Brief, & Guzzo, 1996) behavior holds different perceptions of the organization’s learning culture. Future longitudinal research is needed to investigate the causal impact of the learning culture dimensions on outcomes.

Finally, although our instrument is based on the earlier work of Marsick & Watkins (2003), we acknowledge that their recent publication (Watkins & Marsick, 2023) offers new insights into the practical implementation of learning organizations, particularly in the context of contemporary challenges such as technological changes and global uncertainty. Future research could explore how these new perspectives can be integrated into tools for measuring learning culture and learning organizations.

Despite the general consensus in the literature with regard to the efficient management of constant learning and knowledge as powerful instruments for the maintenance and improvement of an organization’s competitiveness (Jerez-Gomez, Céspedes-Lorente, & Valle-Cabrera, 2005), there is not such a wide consensus in terms of how SMEs can contribute towards a more efficient development of an effective learning capability. Establishing a measurement scale helps reveal the different areas of organizational learning in which managers can act to develop this capability. A relevant implication, therefore, is the approach taken in the activities and relations that need to be present for a firm to be considered a learning organization (Hult & Ferrell, 1997). The scale designed is useful for evaluating more systemic approaches to learning in which the effect of different antecedents of organizational learning can be analyzed. This valid measurement instrument shows that if managers want to improve organizational learning, they can use different means on different levels (individual, team, organization, external environment). Furthermore, the proposed LCS could be implemented as an audit tool. Thus, managers could unveil which organizational learning issues are strong and which are weak. This would provide guidance for targeted action and improvement. In addition, the scale may be useful for analyzing the influence of learning on the firm’s performance, in a search for empirical evidence that supports learning potential as a source of a sustainable competitive advantage (Lei, Slocum, & Pitts, 1999).

The aim of the LCS is also to provide a broader perspective on learning and development within small and medium-sized enterprises (SMEs). In the small and medium-sized business sector, formal courses and training are typically the primary focus when it comes to learning and development, while a culture of learning extends beyond these activities. As a result, SMEs may already be more engaged in learning and development efforts than they realize. The LCS aims to raise SMEs’ awareness of the wider range of learning and development opportunities available and to assist them in leveraging these more effectively, ultimately aiming to foster growth and innovation within their organizations.

The authors confirm that this research has adhered to all applicable ethical standards and guidelines throughout the research process. The study obtained formal and informed consent from all respondents who participated in the survey. The participants were fully informed about the purpose of the research, their rights as respondents and the measures taken to ensure their confidentiality and anonymity. They were also provided with the option to withdraw from the study at any time. Furthermore, the study followed ethical measures in compliance with institutional and international guidelines, including maintaining the confidentiality and anonymity of all respondents; using collected data solely for the purposes outlined in the study; ensuring transparency in data handling and reporting; avoiding any form of coercion or undue influence in securing participation.

Table A1