Recording Business Transactions in Textbook and Receipt Form

An important prerequisite for meeting professional demands upon completion of an apprenticeship is for apprentices to possess the ability to transfer the professional knowledge and skills they acquire in vocational education and training (VET) to various professional situations. This paper presents the empirical results on knowledge and skills transfer from an exploratory study in commercial VET. In a computer-based test environment, 470 commercial apprentices performed business transactions with the same content but with two different forms of presentation. The results indicated that the apprentices did not fully transfer their knowledge and skills within the same context. However, interest in finance and accounting, prior school knowledge and learning opportunities were able to explain their performances. The results suggest that further research on transfer processes that reflect different theoretical approaches is needed.


Introduction
The main objective of vocational education and training (VET) is to provide apprentices with multifarious opportunities to acquire professional competences at different learning venues, which enables them to perform routine activities, deal with new situations and solve various current and potential problems (Illeris, 2004). In Switzerland, this task is promoted in three learning venues: vocational schools, companies and inter-company courses. The normative and shared starting point of all venues is the promotion of apprentices' professional competences. All the same, apprentices have to cope with different priorities in training and work organisations (Eraut, 2004;Schwendimann et al., 2015). Vocational schools take on the task of imparting knowledge and enabling apprentices to reflect on the knowledge they have gained in the workplace in 'protected' settings. In companies, they must conduct activities in real, although varying, situations. Finally, in inter-company courses, apprentices are introduced to the basic skills of the occupation and are prepared for further training in the apprenticing company. Professional competence can only be fully achieved with the ability to link knowledge and skills acquired in these different learning venues and contexts. Therefore, the connections amongst different learning venues play a central role in VET (Aarkrog, 2005;Baartman & de Bruijn, 2011;Kilbrink & Bjurulf, 2013).
Nevertheless, Zitter and Hoeve (2012) noted that experiences and knowledge and skills are often presented in fragmented and uncoordinated ways across learning venues. This deficiency is reflected in the level of research conducted on this topic. Bakker and Akkerman (2019) pointed out that little is known about transitions and connections, whilst Baartman and de Bruijn (2011) spoke of a 'black box' regarding the state of research on the integration of knowledge and skills. Choy et al. (2018) and Wirth (2013a) also concluded that the issue of linking school-based and workplace learning has not yet been adequately addressed. To establish a stronger connection between the learning venues and their contents and activities, vocational schools use both traditional and, as far as possible, realistic learning settings and tasks. This can be seen, for example, in Germany with the learning field concept (Wirth, 2013a) and in Switzerland with a change in the vocational training system to action competence orientation (State Secretariat for Education, Research and Innovation (SERI), 2017).
To date, little research has been conducted in the commercial field to determine how learning venues and their contents and activities relate to one another, whether apprentices are able to transfer the knowledge and skills they have practiced in textbook tasks to more authentic tasks and what factors might influence this transfer. However, such insights are important in assessing whether apprentices can demonstrate their knowledge and skills in similar or different contexts to achieve the common goal of flexibly managing various activities and situations in working life. Therefore, the current paper addresses this question from the perspective of different research strands. In doing so, we present the results of an exploratory study, which investigated whether commercial apprentices in Switzerland can solve related tasks equally well in different forms of presentation (i.e. both in textbook and in more authentic receipt form), which would be an indication of successful knowledge and skills transfer.

Approaches to Knowledge and Skills Transfer
How knowledge and skills learned in (vocational) schools can be applied and linked at work and thus not remain inert has long been a concern of (VET) research (e.g. Hajian, 2019). In their studies on the connection of knowledge and skills, Akkerman and Bakker (2012), as well as Sappa and colleagues (Sappa & Aprea, 2014;Sappa et al., 2018), refer to two different theoretical approaches that are discussed in this context: the comparative and the socio-cultural approaches. The comparative approach, which is also called the 'application approach' (Tanggaard, 2007), is grounded in cognitive-functional models of learning, which are centred on the individual and the processing of information. Concerning the integration process, the focus is on the transfer metaphor, which refers to the acquisition of knowledge and skills applied to a new task, in a new place or in an (un)familiar context. However, it should be noted that there are different definitions and approaches to transfer. 1 These range from a rather narrow understanding of transfer (Thorndike & Woodworth, 1901) to a kind of transfer that prepares for future learning (Bransford & Schwartz, 1999;Schwartz et al., 2005). Schwartz et al. (2005) identified two important factors that influence the transfer of knowledge and skills: individuals' abilities to link prior learning to current and future learning, as well as to react and interpret various learning contexts. Marton (2006) contributed to the transfer discussion by concluding that sameness and difference should be considered. In the literature, different forms of transfer are described, such as high and low road transfers, near and far transfers, as well as positive and negative transfers (Perkins & Salomon, 1992). Barnett and Ceci (2002), for example, defined 'near transfer' as the acquisition and application of knowledge and skills in a similar context, whilst 'far transfer' referred to the acquisition and application of knowledge and skills in different contexts.
To enable transfer, defined as 'onetime and one-sided transitions' (Akkerman & Bakker, 2011, p. 136), Akkerman and Bakker (2012) as well as Sappa et al. (2018) argue for closing or reducing differences amongst the learning venues by imparting knowledge and skills in a similar way at all places of learning. This means that curricula should be aligned amongst the learning venues (Sappa et al., 2018) and that vocational schools should establish links to the real world by relating to authentic problem situations (Mandl et al., 1992). However, the comparative approach has been criticised by Tanggaard (2007), amongst others, for being oversimplified, e.g. assuming that knowledge and skills are not affected by transfer and application.
In response to this objection, the sociocultural approach, also known as the relational approach, focuses on how learning venues relate to one another with respect to learning. Learning and, therefore, learning transfer, cannot take place independently of context (e.g. where apprentices learn and work) and social interaction (e.g. with whom they learn and work). The theoretical basis for this approach is constituted, amongst others, by the third generation of cultural historical activity theory (Engeström, 2004) and the situated learning theory (Lave & Wenger, 1991;Wenger, 1998). As an alternative to the aforementioned more traditional transfer metaphor, the so-called boundary crossing metaphor is proposed (Säljö, 2003;Tuomi-Gröhn et al., 2003), which, according to Akkerman and Bakker (2011, p. 136), stands for 'ongoing, two-sided actions and interactions between contexts.' Against this background, the sociocultural approach does not pursue an approximation process amongst the venues, but rather sees congruencies and divergences as a decisive factor in learning and development processes. For learning and teaching in VET, this implies that apprentices' experiences and learning at the different learning venues should be related to one another and reflected upon. This process includes active communication amongst all parties involved (Sappa et al., 2018).
The two theoretical approaches indicate that different perspectives can or should be considered when evaluating apprentices' knowledge and skills transfer. However, similar to Sfard's (1998) article on learning, the point is not to pit one approach against another. Rather, it is a matter of advocating one approach based on a particular research focus without negating the other or even attempting to combine them. In the present study, we follow the comparative approach and understand transfer as the transition of acquired knowledge and skills across various contexts, according to Barnett and Ceci (2002). In this way, we aim to investigate the extent to which Swiss commercial apprentices can apply knowledge and skills from one form of presentation to another within the same context, which can be understood as a form of near transfer.

Measurement of Knowledge and Skills and their Transfer
It can be assumed that knowledge and skills are integral components of competence (Biemans et al., 2004;Blömeke et al., 2015). By competence, we mean the application of knowledge and skills as well as affective-motivational elements (dispositions), moderated by situation-specific skills (e.g. perception, interpretation), in reallife situations (performance) (Blömeke et al., 2015). Measuring this performance allows conclusions to be drawn about competence and, therefore, knowledge and skills. Similar to competence, transfer is not directly visible or measurable; instead, it can only be derived from the evaluation of professional actions or knowledge and skills.
Different test modes (e.g. paper-pencil, simulations, observations) and different task formats (e.g. open text, multiple choice, rubrics) have been applied to measure knowledge and skills. Since the shift to competency-based curricula in VET and specifically in the commercial field, paper-pencil tests (Böhner, 2005;Rosendahl & Straka, 2011;Seeber, 2007;Stegemann et al., 2015) and, from at least 2009 onwards, computer-based tests (see projects in the ASCOT line of funding, Beck et al., 2016) have been used to measure knowledge and skills. The use of computer-based test formats is associated with the assumption of testing that is practical and close to reality, as well as with the implementation of tasks that are similar to actual workplace situations and potentially have a higher action orientation (Jude & Wirth, 2007;Sangmeister, 2017;Seeber, 2016). However, the goals of the aforementioned projects were not to 'measure' transfer but to assess applied professional competence as a whole. One piece of research from the commercial field that could be linked to transfer considerations is the work of Sangmeister (2017). To examine the influence of testing modes, she investigated whether apprentices solved test items better in a paper-pencil test or in a simulation (preferably in the same form of presentation, enabled by screenshots of the simulation). However, she did not administer identical tasks in either mode. Sangmeister (2017) concluded that (domain-)specific items were handled better in the computer-based form, as they correspond more to authentic situations, whereas 'general items' were solved better in the paper-pencil form.
Studies that particularly focused on the transfer concept were carried out in the 1990s. Some of these concluded that in various fields, such as everyday mathematics (Lave, 1988;Saxe, 1988) or medicine (Boshuizen & Schmidt, 1992;Gräsel et al., 1993Gräsel et al., , 1994, existing (prior) knowledge could not be applied to new situations and thus remained 'inert'. By contrast, VET research in the commercial domain, which quantitatively investigated the transfer of knowledge and skills (acquired at vocational school) into concrete professional actions, remains limited. What the few available studies have in common is that they recognise the existence of problems with knowledge (and skills) transfer. For example, Mandl et al. (1992) demonstrated that commercial apprentices were unable to utilise their declarative knowledge in a simulation of a business process. Holtsch et al. (2017) came to a similar conclusion in their analysis of whether commercial apprentices were able to use basic concepts of mathematics and German in commercial situations. Maag (2019) investigated learning processes (which he called 'dual learning' 2 ) across learning venues and placed particular emphasis on the influence of prior knowledge. He concluded that there is a bivariate correlation between prior knowledge and learning success in the field of financial accounting. However, he drew this conclusion, even though he was only able to demonstrate in one-third of the learning processes recorded that the apprentices used their prior knowledge from one learning venue for new experiences in the other learning venue. In summary, following Jackson et al. (2019), whilst knowledge and skills transfer is often assumed, specific empirical evidence is lacking. This can also be noted for the commercial domain.
The following section will explore in more depth the factors from different learning venues that can explain and promote the concept of transfer.

Factors Promoting Transfer and Explaining Learners' Performance in Competence Tests
An overview of factors explaining the 'integration of knowledge, skills and attitudes' was provided in a review study by Schaap et al. (2012). 3 The authors concluded that prior knowledge is relevant to integration in terms of the processing of new information. Moreover, they found that students (known as 'apprentices' in the current work) perceived learning in workplaces (sometimes) as overtaxing, that it was necessary for them to act in various situations and contexts and that it was essential to reflect on divergences and similarities amongst different contexts.
Regarding the factors explaining performance in competence tests and transfer, several key findings are currently available. Opportunities to learn (OTL) in the richest possible situations and contexts have been shown to be relevant to effective learning (Billett, 2001;Fuller & Unwin, 2004). As far as company characteristics are concerned, the results on the relevance of company size are ambivalent. Hayton et al. (1996) andForster-Heinzer et al. (2020) reported different forms of OTL, depending on company size. Virtanen et al. (2014), in contrast, could not present empirical evidence of a direct influence of company size on learning.
Regarding individual characteristics, prior knowledge has been proven to influence learning success (Billett, 2018;Maag, 2019). In particular, Maag (2019) found that prior knowledge of one learning venue was a good predictor of learning success at the another learning venue (r = 0.40). However, prior knowledge at vocational schools plays a more important role in workplace learning than vice versa. Past studies have provided empirical evidence suggesting that cognitive abilities are relevant to the development of vocational knowledge and skills (e.g. Atik & Nickolaus, 2016;Helm, 2015;Rosendahl & Straka, 2011;Sticca et al., 2018). Moreover, interest in the learning of occupational knowledge and skills may be a key factor in learning transfer or integration (Billett, 2015;Wild, 2011). Another important factor for successful knowledge and skills transfer seems to be the perception of learning and integration at the learning venues. Diepold (1996), Wirth (2013aWirth ( , 2013bWirth ( , 2015 and Sappa and Aprea (2014) concluded that in many cases, apprentices do not make connections between learning in vocational schools and in companies, which can lead to difficulties in the completion of knowledge and skills transfer. As for laboratory technician apprentices, such problems could be explained with reference to the findings of Akkerman and Bakker (2012, p. 167), who observed that experiences at work were 'technology-mediated, scripted and socially distributed', thus leading to implicitness and fragmentation. This is problematic because apprentices who were able to better perceive the integration between learning venues showed better self-assessed workplace learning outcomes (understood as skills learning and identity development), according to Virtanen et al. (2014). Jackson et al. (2019, p. 462) recognised a large consensus in the research literature that (the already mentioned) individual and company characteristics, as well as learning programme characteristics, play a crucial role in knowledge and skills transfer. For example, school-work alignment has been mentioned in many studies as a relevant factor in the transfer of learning (Kilbrink & Bjurulf, 2013;Mulder et al., 2015;Renta Davids et al., 2017), especially because the relevance of prior learning is emphasised. Kilbrink and Bjurulf (2013) also stress other factors, including communication between vocational schoolteachers and trainers in the company, reflection and financial resources.

Interim Summary
Despite these theoretical and empirical results, the findings regarding apprentices' knowledge and skills transfer are still deficient in at least three aspects. First, according to Schaap et al. (2012), it remains empirically vague how knowledge, skills and attitudes are combined and transferred into professional competence. To explore this, it would be of interest to initially investigate how apprentices of a large and representative sample solve related tasks on the same learning content that require knowledge and skills transfer. Second, relevant factors, such as the perception of learning and integration at the learning venues, have been examined, but mostly from a qualitative perspective. Therefore, it would be informative to provide results from a larger sample on these factors. Third, relevant factors, such as apprentices' interests or learning experiences, can explain knowledge and skills transfer. Nevertheless, research on the interplay between knowledge and skills transfer (measured in terms of concrete professional actions) and relevant factors has rarely been conducted using quantitative methods.

Research Questions
Based on the findings and the desiderata presented above, we aim to address the following research questions:

How do commercial apprentices deal with related tasks in different forms of presentation?
We assume that (near) transfer is successful when related tasks in different forms of presentation can be solved identically.

How do commercial apprentices assess and describe factors that relate to the transfer of knowledge and skills?
Several factors can be assumed to explain the extent of knowledge and skills transfer. Therefore, first, we aim to describe the commercial apprentices' experiences (finance and accounting as a typical activity), their individual characteristics (e.g. interest in finance and accounting, perception of opportunities to apply financial knowledge in the workplace) and company characteristics (company size) from a quantitative perspective.

To what extent do factors related to knowledge and skills transfer explain apprentices' performance on related tasks in different forms of presentation?
Building on Research Question 2, we assume that the quantitative levels of these factors have an impact on the performance of related tasks presented in different forms.

Study Context
Commercial apprenticeship is the most frequently chosen VET programme in Switzerland. Approximately 14,000 commercial apprentices per cohort receive threeyear training in one of 21 different commercial branches (Swiss Conference of Commercial Training and Examination Branches (SKKAB), 2020). 4 In this system, apprentices learn one to two days a week at vocational schools, spend three to four days in their workplaces in companies and regularly attend branch-specific courses.
This exploratory study on transfer is a 'spin-off product' of the longitudinal research project Learning and Instruction for Commercial Apprentices (LINCA), which investigated teaching and learning processes in basic commercial training in Switzerland. 5 Amongst other things, commercial apprentices were asked throughout their entire vocational training about different competence areas, such as commercial knowledge and skills. To survey as many apprentices as possible at the same time and to enable a controlled test situation, the apprentices were surveyed in vocational schools only.

Measurement of Knowledge and Skills Transfer
To measure the apprentices' knowledge and skills transfer, we used a selection of items from the computer-based test instrument simuLINCA, which was initially devised to measure the commercial knowledge and skills of Swiss commercial apprentices (Rohr-Mentele & Forster-Heinzer, 2021). In the test environment, the apprentices took on the role of employees at the fictional company LINCA. They first had to solve working tasks (e.g. processing of a request for quotation), which measured their commercial skills, and were then expected to respond to items in a mostly closed format to assess their commercial knowledge. The test comprised 76 items, 33 of which were related to knowledge and 43 to skills. The reliability scores of the test were 0.78 for the knowledge dimension and 0.86 for the skills dimension (Rohr-Mentele & Forster-HeinzerAuthors, 2021).
To investigate the extent to which apprentices could make a transfer between related tasks in different forms of presentation, we opted for the competence area 'executing financial processes' because this is promoted in both vocational schools and in companies (relevance for future work). In vocational schools, apprentices must solve textbook-and receipt-based tasks in accounting. In the final examinations given at the end of the apprenticeship, they are required to solve both types of task formats. However, the textbook form seems to dominate both vocational school instruction and final examinations. 6 In the company, there are also learning objectives for accounting, but the apprentices are usually required to use software, such as that from SAP. Then, we created both receipt-and textbook-based tasks on the same learning content to map typical learning opportunities in commercial VET and measure transfer.
In the test, the apprentices must first solve the receipt-based accounting task ( Fig. 1). At the beginning of the test, simuLINCA, they were required to perform a price calculation for a beverage. After that, they were requested to record transactions concerning this product. Therefore, the trainer, Nicole Müller, sent an email  Table 1) and asked the apprentices to record the transaction based on two receipts (order and bank) and the charts of accounts. To solve this task, the apprentices must first find out which and whose financial activities took place (i.e. who bought which product at which price (cost of goods) and with which VAT rate). To do this, they must analyse the two documents in terms of totals and dates and consider multiple accounts when recording financial activities. In the competence level model for external accounting proposed by Guggemos and Schönlein (2015), the task would correspond to a higher level in the dimensions of 'information procurement' and 'information processing'. In vocational schools, apprentices are used to recording receipt-based tasks, but these are almost never simulated on a computer. In the company, apprentices receive receipts, which are then entered into a corresponding system. In our computer-based test, we partially simulated a 'SAP system', so the task did not quite authentically match the real situation in the company. However, the apprentices in the pilot studies found the task and setting to be quite realistic for the company.
After working on the receipt-based task, the apprentices recorded two business transactions in textbook form that dealt with the same content; thus, it had the same number of recordings as the simulated accounting task (Fig. 2). The task is cognitively less demanding, as it is clearly structured and the necessary information is fully provided. This form is only authentic for vocational schools.
The description and main differences between the two forms of presentation are shown in Table 1. For both forms, five accounting records had to be entered in the columns pertaining to the following: (a) the debit account, (b) the credit account and (c) the amount of money with the aid of a chart of the accounts. The five accounting  Table 2) records included the following: 1) expenditure on goods, 2) value-added tax (VAT), 3) (bank) payment, 4) cash discount and 5) VAT on cash discount. The correct solution for the textbook task, along with the partial correct examples of answers from the apprentices, can be found in Table 2.

Measurement of Factors Explaining Performance and/or Transfer
Based on the considerations stated in the theoretical part of the paper, we included variables that can potentially explain the extent of apprentices' knowledge and skills transfer.   The apprentices rated the following: (a) the extent to which they perceived cooperation between vocational schools and companies, using a Likert-type scale ranging from 1 'not at all existing' to 6 'existing to a very high degree', and (b) the extent to which they applied their school knowledge about finance and accounting in their workplace from 1 'never applied' to 6 'very often applied'. (4) Cognitive abilities and prior knowledge: (a) Cognitive abilities were measured by means of the revised short form of the Cognitive Ability Test (CAT) for 4 th to 12 th grade (Heller & Perleth, 2000), which included the subtests V3 (word analogies, 20 items), Q2 (series of numbers, 20 items) and N2 (character analogies, 25 items). 7 (b) In the simulation, the apprentices had to solve three independent accounting multiple-choice tasks (e.g. selection of an accounting record for a described business transaction). The multiple-choice tasks were regarded as representing their prior school knowledge about accounting. (5) Interest in learning in a specific occupational domain: Using a Likert-type scale ranging from 1 'not at all interested' to 6 'extremely interested', the apprentices had to indicate how interested they were (a) in finance and accounting (school) and (b) in executing financial processes (company).

Sample
In spring 2015 (t 1 ), a total of 1,365 commercial apprentices from 82 commercial schools in the German-speaking part of Switzerland were surveyed in their third and final year of training. Amongst the apprentices, 36.8% were male, and the mean age was 19.1 years (SD = 1.5). Approximately three months later, in early summer 2015 (t 2 ), the same apprentices participated again, including 1,214 commercial apprentices from 76 commercial schools. Apprentices usually prepare for their final examinations between spring and summer. Therefore, this short period can be assumed to be important for the acquisition of commercial knowledge and skills. An overview of how the apprentices completed the accounting tasks at t 1 and t 2 is provided in Table 3. 8 In the following analyses, we included only 470 apprentices (38.7%) who had completed the two accounting tasks (textbook and receipt form) at both measurement points.

Coding of answers:
The categories for the coding of the five accounting records were developed before data collection based on the correct solution and assumptions about misconceptions (e.g. confusion between net and gross method). When the data collection was completed, new coding categories were added after the initial data analysis (e.g. frequencies of answers). The apprentices who had dealt with all three components (debit account, credit account and amount of money) correctly were given a score of two points (fully correct). Those who had included two components received one point (partially correct). Approximately half of the cases were double-coded. The agreement of the two trained coders reached Cohen's kappa ranging from 0.814 to 0.990 (t 1 and t 2 ) for the receipt form and from 0.916 to 0.983 (t 1 and t 2 ) for the textbook form, which corresponds to an almost perfect value, according to Landis and Koch (1977). All subsequent analyses were conducted using IBM SPSS Statistic software (Version 26).

Research Question 1 (RQ1):
To answer this question, we calculated the frequencies and mean values of the two forms of accounting records (points scored for each record and sum score of all records) for both measurement points. Then, we analysed the significant changes over time (t 1 -t 2 ) using a paired sample t-test and effect sizes according to Cohen (1992). Furthermore, we calculated the Pearson correlation coefficients of the test points (sum score of both forms) and computed a difference variable between the two forms.
Research Question 2 (RQ2): To answer this question, we drew on the descriptive results concerning the factors explaining knowledge and skills transfer. In the longitudinal project LINCA, we collected data on most of the factors multiple times, including testing the conjecture of the extent to which the factors remained stable over time. Whether changes over time (t 1 -t 2 ) were significant was checked by a t-test for paired samples.
Research question 3 (RQ3): Before we performed the regression analyses, we checked whether the requirements for calculation (linearity, no outliers, independence of the residuals, no multicollinearity, homoscedasticity and normal distribution of residuals) were given. With respect to outliers, leverage values (Igo, 2010;Velleman & Welsch, 1981), 9 Cook distance (Cook, 1977) and studentized deleted residuals (-/ + 3 SD) were tested; cases were excluded from the analyses if at least two procedures detected the same outlier. This was, however, never the case. Regarding the independence of residuals, both graphical (scatter plot) and the Durbin-Watson test were used. Regarding multicollinearity, we checked the tolerance value (T), which should be higher than 0.10 or, even better, higher than 0.40, and the values proved to be unproblematic for all calculated models (Cohen et al., 2003). Homoscedasticity was checked graphically (scatter plot). Basically, it can be stated that the requirements for all models are not ideal but acceptably fulfilled.
In the theoretical part of the paper, we assume that experience, company characteristics and individual characteristics are relevant to a successful knowledge and skills transfer. Therefore, we generated an explanatory model for each of the three aspects and one model that included all factors. We performed these calculations for the first measurement point because all factors were only available for t 1.
10 For all models, the dependent variables were either the sum score of the five accounting records in textbook form or the sum score of the five accounting records in receipt form. The independent variables, which were all interval-scaled or dummy variables, were included or excluded, depending on the model. Model 1 and Model 2 were simple linear regressions. The first model, 'Experience', (1a/b) included the variable 'Finance and accounting as a typical activity' in the participants' workplace. The second model, 'Company characteristics', (2a/b) included the variable 'Company size', which had previously been ordinally scaled and, therefore, needed to be recoded (0 = '0-50 employees' and 1 = 'more than 50 employees'). As we assumed that all factors were equally relevant, we chose the method of simultaneous multiple linear regression for the other two models. The third model, 'Individual characteristics', (3a/b) related to the perception of cooperation, the perception of opportunities to apply knowledge about finance and accounting, cognitive ability, multiple-choice tasks on accounting, as well as interest in finance and accounting (school and company). In the fourth model (4a/b), all factors were introduced.

Analyses Concerning the Performance of Accounting Tasks (RQ1)
One aim of this study was to determine whether apprentices performed similarly on an accounting task in two forms of presentation. Referring to the percentage values (Fig. 3), the results showed that four of the five accounting records were slightly better when the task was presented in textbook form, both at t 1 and t 2 . The first accounting record concerning the expenditure on goods was the only record that proved to be better when the receipt form had been given, as indicated by the mean scores (Table 4) and the number of correct answers (Fig. 3). As for accounting records 2 (VAT) and 3 (payment), the apprentices scored more points with the textbook form than with the receipt form, both in terms of mean values, correct answers and partially correct answers. They performed poorly on accounting records 4 (cash discount) and 5 (VAT cash discount) in the receipt form but did much better in the textbook form, especially at t 2 . In the receipt form, depending on the task and the measurement point, only 0.2% to 4.8% of the apprentices achieved the full score,   (Table 4, SUM) ranged from 2.92 to 5.35 points out of a possible 10. Overall, the discrepancy between the two forms tended to increase towards t 2 in accounting records 1, 4 and 5.
As reported in Table 4, in all accounting records (receipt and textbook), except record 3 in the receipt form, the apprentices improved their performance significantly from t 1 to t 2 , with effect sizes ranging from Cohen's d = 0.10 to 0.64. Particularly regarding accounting records 4 and 5 in textbook form, they achieved significantly better results at t 2 , with an effect size of d = 0.59 and d = 0.64, corresponding to a medium effect.
We assumed that similar performances on both receipt and textbook form (measured in points) would indicate the extent of successful knowledge and skills transfers amongst the apprentices. The next step, therefore, was to determine the extent to which the two tasks in different forms of presentation correlated with each other.
The sum scores of the completed accounting records correlated at both measurement points to a weak to medium degree. This means that the apprentices who performed better on one form also scored higher in the other form with a medium effect (t 1 : r = 0.419, p < 0.01; t 2 : r = 0.396, p < 0.01).
To understand the correlation results in more detail, we calculated the difference points between the sum scores of the textbook form and the receipt form for both measurement points (Fig. 4, x-axis). As 2 points are awarded for each of the five correctly solved accounting records, 10 points per form can be achieved. For example, 0 difference points meant that both forms were dealt with equally well. By contrast, + 10 (-10) difference points indicated that all accounting records in the textbook form (receipt form) were performed correctly, whereas the apprentices received 0 points in the receipt form (textbook form).

Fig. 4 Distribution of the Difference Points
At t 1 , 20.4% of the apprentices achieved identical scores, meaning 0 'difference points' between the two forms. However, 6.8% of the apprentices scored 0 points, and 2.5% of the apprentices scored 1 point on both tasks. Around 26.4% of the apprentices solved the tasks within a deviation point. Yet, 5.5% of these apprentices scored 0 points on the textbook form and 1 point on the receipt form or vice versa. In other words, 14.8% of the apprentices solved the tasks (nearly) identically but (mostly) incorrectly. The other half of the apprentices showed differences in how successfully they dealt with the tasks; 12.8% of them had 5 or more deviations in points.
At t 2 , 18.7% of the apprentices achieved identical scores. However, 4.5% of the apprentices scored 0 points, and 1.1% scored 1 point on both tasks. Around 22.3% of the apprentices solved the tasks within a deviation point. Yet, 2.8% of the apprentices scored 0 points on the textbook form and 1 point on the receipt form or vice versa. The increase in deviation at t 2 , amongst others, amounted for 19.8% of the apprentices to 5 or more deviations in points. Notably, there was a slight shift in the values into the 'plus area'. A paired sample t-test confirmed that the results differed significantly from t 1 to t 2 (T = -2.333, df = 469, p < 0.05).

Descriptive Results for Relevant Factors (RQ2)
Table 5 provides the descriptive results for the relevant factors that can be assumed to contribute to explaining the extent of knowledge and skills transfer. Approximately 30% of the apprentices stated that finance and accounting is one of their typical tasks in the workplace. The surveyed apprentices were predominantly employed in medium-sized (50-249 employees) and large companies (250 and more employees), whereas 28.6% of them worked in companies with less than 50 employees.
The apprentices perceived the cooperation between vocational schools and companies as partially existing (M = 3.60, SD = 1.16) and reported that they could partly apply the knowledge about finance and accounting they obtained in school to their tasks in their respective workplaces (M = 3.34, SD = 1.44).
In the cognitive ability test, the apprentices scored an average of 41.06 out of a possible 65 points. As for the multiple-choice tasks on accounting, the apprentices solved, on average, about two of three tasks correctly (M t1 = 1.83, SD t1 = 0.86; M t2 = 2.02, SD t2 = 0.77). From t 1 to t 2 , the apprentices improved their performance significantly in the accounting multiple-choice tasks. Moreover, they reported being interested in finance and accounting (M t1 = 3.96, SD t1 = 1.35; M t2 = 3.79, SD t2 = 1.46), which proved to be slightly higher in school than in the company (M t1 = 3.80, SD t1 = 1.24; M t2 = 3.57, SD t2 = 1.27). However, interest decreased significantly from t 1 to t 2 .

The Relationship Between Performance and Relevant Factors (RQ3)
Regression analyses were calculated to predict the apprentices' performance in accounting tasks based on relevant factors. The results are displayed in Table 6. The regression analyses of the first two models, 'Experience' (Model 1) and 'Company   Characteristics' (Model 2), showed that both variables, 'Finance and accounting is a typical workplace activity' and 'Company size', did not provide significant explanatory power on their own, particularly regarding the variance in the sum score of the accounting records in either form of presentation. Model 3, 'Individual characteristics', explained 12.1% of the variance in the textbook form and 11.7% of the variance in the receipt form. In the textbook form, 'Accounting multiple-choice tasks' (B = 0.915, p < 0.000) and 'School interest in finance and accounting' (B = 0.320, p = 0.006) contributed significantly to the explanation of the performance. By contrast, in the receipt form, only 'Accounting multiple-choice tasks' proved to be significant (B = 0.778, p < 0.000). This indicates that accounting multiple-choice tasks have a slightly higher effect on the textbook form than on the receipt form.
Compared to Model 3, Model 4, which included all factors, led to a slightly higher variance explanation of 13.2% in the textbook form and to a similar variance explanation of 11.8% in the receipt form. This corresponds to a moderate effect (Cohen et al., 2003). When 'Accounting multiple-choice tasks' and 'Interest in finance and accounting at school' increased by one scale point, the sum score of the textbook form (Model 4a) increased by 0.946 and 0.310, respectively. By contrast, when the scale value 'Finance and accounting is a typical activity in the workplace' increased by one point, the sum score decreased by 0.756. This represents a change compared to Models 3a and 1a. In Model 4b (receipt form), both interest and finance and accounting as a typical activity in the workplace were not significant. When 'Accounting multiple-choice tasks' increased by one scale point, the sum score of the receipt form increased by 0.803. This indicates that, similar to Model 3, accounting multiple-choice tasks is a slightly better predictor of the textbook form than of the receipt form.

Discussion and Conclusion
Knowledge and skills transfer is a topic that is being researched from several perspectives in VET. In this regard, various concepts of transfer are discussed, such as near transfer (i.e. the acquisition and application of knowledge and skills in a similar context), which we have considered in more detail. However, little is known about the extent to which apprentices transfer their knowledge and skills within or across learning venues and contexts (Bakker & Akkerman, 2019;Choy et al., 2018). The current paper addressed this research gap and presented results on the measurement of knowledge and skills transfer in the context of the recording of business transactions.

Interpretation of the Results
Despite similar content, the apprentices performed differently on the accounting task depending on the form in which the task was presented (RQ1). For some apprentices, the differences in performance between the forms amounted to nearly 30%. 11 One-half of the apprentices solved the two tasks (within a deviation point) equally well, which would-according to our interpretation-indicate that transfer had been successful. However, at t 1 , many apprentices scored 0 points on both forms, indicating that the tasks were solved identically, but not in the intended way. The other half of the apprentices showed (major) differences in dealing with the two forms. As we assume that near transfer was successful if both forms of accounting tasks were handled equally well, knowledge and skills transfer seemed to have been only partially accomplished by the apprentices. This means that, without further assistance, the majority of the apprentices would be unable to conduct (routine) accounting workplace activities to a similar extent or with similar content at the end of their commercial VET programmes.
From a subject-specific pedagogy perspective regarding the recording of business transactions, it could be that the apprentices performed better on the textbook form of accounting record 2 (VAT) than on the receipt form, because VAT was explicitly mentioned in the former but not in the latter. This explanation also applies to accounting records 4 and 5, because the step of the solution concerning (VAT on) cash discount was not explicitly mentioned in the receipt form. In other words, if the task is more open and complex, as is the case in reality, it would be more difficult for apprentices to transfer their knowledge and skills and solve the task at hand. This could also indicate that apprentices have deficits in complex problem-solving skills. However, the apprentices also had difficulty solving the textbook form.
From a teaching-learning perspective, the results could be explained by the assumption that apprentices are generally more used to learning and teaching in textbook form and that they might have perceived the simulated accounting task in receipt form as not so authentic. Although their overall performance improved from t 1 to t 2 , they still performed better in the textbook form. This could be due to the training effects at school that took place between t 1 and t 2 , particularly because the final examinations were forthcoming, and textbook-based tasks were more often implemented than receipt-based tasks. However, this assumption would need to be tested with data on classroom task analyses, for instance.
The second research question (RQ2) dealt with the apprentices' descriptions of relevant factors that promoted transfer, such as opportunities to learn, as discussed in the literature. About 30% of the apprentices stated that finance and accounting tasks were amongst their three main activities in their respective workplaces. At first glance, 30% seems to be a relatively low percentage for commercial VET compared to the curriculum. This result might be attributed to the increasing extent of digitalisation, such as the outsourcing of financial processes like corporate accounting (e.g. Iten et al., 2016). The rather low scores of the scales concerning the perception of cooperation between the learning venues and the application of accounting knowledge point to an understanding of learning as a distinct experience depending on the venue, which is consistent with the findings of Diepold (1996), Wirth (2013aWirth ( , 2013bWirth ( , 2015 and Sappa and Aprea (2014). However, it would have been desirable if both values had been higher, as the results from Virtanen et al. (2014) indicate that workplace outcomes also increase with a more pronounced perception of cooperation. On average, the apprentices correctly solved two of the three multiple-choice tasks on accounting, indicating that they possessed basic knowledge and probably solved similar tasks at vocational school. Furthermore, they reported interest in finance and accounting, which proved to be slightly higher in school (even though measured with one item) than in the company. The finding that their interest decreased during the apprenticeship is in line with previous results on the development of interest in VET (Forster-Heinzer et al., 2020;Krapp & Lewalter, 2001).
Relevant factors can explain-at least to a limited extent-the overall performance of the apprentices in accounting tasks, which we assume to be indicative of their knowledge and skills transfer (RQ3). The most comprehensive regression model (4a/b), which included all factors, proved to be the one with the highest explanation of variance (14.8%/13.5%). Regardless of whether the textbook or the receipt form is emphasised, the fourth regression model in particular shows a similar pattern: no effects of company size, cognitive ability, interest in company finance and accounting and perception of cooperation. In particular, the finding that perception had no influence may be surprising. This could be attributed to the fact that we included only one item, which seemed to be too global for the purpose of measuring knowledge and skills transfer. The lack of relevance of interest in the company's finance and accounting could indicate that the receipt-based task is not authentic enough (e.g. in terms of the materials used (software) or the relevance of VAT entries). This assumption would have to be examined in future studies applying, for example, think-aloud analyses.
For both forms, prior school knowledge, as measured by multiple-choice tasks on accounting, was the strongest predictor of performance. This result is in line with Maag's (2019) findings regarding the relevance of prior knowledge. Moreover, our empirical findings on interest in finance and accounting at school also confirm previous research results on the relevance of interest for knowledge and skills and vice versa (Billett, 2018;Schaap et al., 2012). This implies that apart from facilitating experiences at learning venues, interest should also be promoted and maintained in school settings.
A difference between the two forms of presentation manifested itself with respect to the textbook form, namely, as a negative (weakly significant) effect when finance and accounting had to be applied in the workplace. This does not correspond to previous research findings that emphasised the relevance of opportunities to learn (Billett, 2018). Rather, our findings seem to indicate that company practice is at odds with the textbook approach in accounting in vocational schools. This could be explained by the fact that, at present, the performance of financial activities in the workplace is not examined in depth due to technology-mediated and abstract work processes (Akkerman, & Bakker, 2012;Wirth, 2013a). Another assumption is that apprentices attribute little realism to the tasks in textbooks and therefore attach little importance to them later in their professional careers, which is reflected in their poorer performance. Assuming a significance level of 10%, another difference emerged regarding the receipt form: the perception of opportunities to apply financial activities in the workplace promoted the apprentices' ability to deal successfully with the receipt form. These two differences imply that practice in company matters, as suggested in the sociocultural approach. On the one hand, company practice seems to counteract the 'traditional' practice of teaching accounting in schools (textbook-based task); on the other hand, it promotes the solving of the receipt-based task. Even in the context of digitalisation and computerised accounting processes, it is important to be able to trace accounts and correctly interpret receipts and any erroneous messages.

Limitations of the Study
Our findings should be considered in light of two main limitations regarding the measurement instruments and the lack of in-depth analyses of learning processes in vocational schools and companies.
First, in measuring knowledge and skills transfer, only the accounting tasks were implemented in the instrument, simuLINCA. It would be crucial to develop and apply further commercial tasks to assess and describe apprentices' commercial knowledge and skills transfer in more detail and from a holistic perspective. There are also limitations regarding the instruments used to measure the factors explaining such a transfer. For example, we administered only one item each on perception of school and workplace cooperation and on perception of opportunities to apply financial accounting knowledge in the workplace. Furthermore, data on finance and accounting as a typical activity in the workplace, the perception of school and workplace cooperation and the perception of opportunities to apply finance and accounting knowledge in the workplace were only gathered at t 1 . Whilst interpreting the data, it must also be noted that the comparative approach alone is not sufficient to explain the results. Factors of the sociocultural approach would also have to be included in a follow-up study. This leads to a methodological question of how and which factors would have to be considered and measured in a combination of comparative and sociocultural approaches. The theoretical part of the paper points to scales, such as 'linking prior learning', 'reacting to and interpreting learning contexts' and 'amount of alignment' for the comparative approach. To account for the sociocultural approach, scales such as 'communication between parties', 'recognizing and reflecting on congruencies and divergences' and 'company practices' would have to be used. A common factor in both approaches is that learning opportunities (presented in different situations and contexts) must be considered in detail. Thus, the extent to which Renta Davids et al.'s (2017) questionnaire on individual, educational and workplace factors can reflect both approaches and could be used in a follow-up study is worthy of further exploration.
Second, our results demonstrated the extent to which the apprentices transferred their knowledge and skills in terms of the recording of accounting tasks. However, to optimise learning and teaching, it would also be useful to obtain in-depth knowledge about apprentices' learning processes. Related to this, Baartman and de Bruijn (2011) provided a theoretical basis for integrative learning processes. Yet, their theory must be confirmed empirically.

Conclusion
This paper addressed the question on Swiss commercial apprentices' ability to transfer knowledge and skills. Furthermore, we examined which relevant factors can explain apprentices' general performance in competence tests and might foster knowledge and skills transfer. The results showed that accounting tasks in textbook and receipt forms were solved differently, even though the recording of the business transactions had the same content and concept. This indicates an insufficient transfer of knowledge and skills. Therefore, we conclude that what is learnt in one form and/ or context in vocational schools cannot be fully applied to a new form and/or context, especially in a company setting.
At the same time, experience and individual characteristics can explain the extent of transfer to a limited degree. In particular, the relevance of prior school knowledge and interest was empirically demonstrated in our study, but not company size or the apprentices' perceptions of opportunities to apply financial accounting knowledge in the workplace. We also concluded that additional factors, such as communication between parties, should be included in a follow-up study.
Meanwhile, there is a need for further research on the conceptualisation and effective implementation of integrative learning processes (Schaap et al., 2012). It is important to know where and how, or under what conditions, apprentices acquire their professional competences and, even more so, how the transfer of knowledge and skills takes place within and between learning venues. Such learning processes should also be empirically measured and made 'visible', such as through documentations, portfolios or videos of tasks performed in a workplace. In fact, these could provide information about the effective design of teaching and learning processes at learning venues.