| Summary |
General overview and research goals:
Compared to the body of research focusing on teaching and learning in grade school, research on quality in VET is still scarce. Also diagnosis of competencies in mechanical training is currently not based on empirically founded instruments and results.
This paper is a contribution to this and asks two main questions:
• How can we measure competencies in the field of mechanical training and how can we do that economically?
• How can we structure different aspects of competencies?
To answer these questions, we first define competence and suggest a theoretical model for structuring competence. Secondly, we exemplify how we can approach the problem of competence diagnosis in an economical, reliable and valid way, and how we can test our theoretical model empirically.
Theoretical Background:
Generally, in the vocational educational literature "competence" is defined as following: Someone who is competent is authorized to act and takes over responsibility for him/herself and other persons. The person’s activities are motivated by an intention, a goal or a purpose and are guided by action principles, values, standards and rules. Someone who is competent acts successfully, reflectively and consciously. We consider competence as a bundle of physical and mental abilities enabling a person to solve tasks or problems. Four competence-classes can be defined: “Specialized competencies”, “social competencies”, “cognitive competencies”, and “personal competencies” (Bader, 2002; Frey, Jäger & Renold, 2005).
Specialised (spec), social (socc), cognitive (cogc) and personal competence-classes (perc) are not independent from one another. On the contrary, they form networks (Frey, 2004). According to Roth (1971) and Frey (2006), a theoretical model of competencies in vocational education could look like this:
appendix 1, figure 1:
The competence class “specialized competencies” is correlated with “cognitive competencies”. Both of them have an effect on “social competencies”. Furthermore “specialized competencies” and “social competencies” have an effect on “personal competencies”, which is the end of the effect-series. This model stresses the important role of “specialized competencies” (see Frey, 1999).
Methodology:
In the context of the dual system of vocational education and training (VET) in Switzerland, the longitudinal study QuWibB (Quality of Vocational Training in Enterprises, see http://www.quwibb.info) is investigating conditions and training practices at the workplace, i.e. in firms offering apprenticeships in the area of mechanical training. In particular, we are questioning apprentices being trained as polymechanics about their training at work. They go through a four-year long training program that involves knowledge and use of high technology tools and machines.
Beside the other goals which are pursued in the project, we need to diagnose competencies in a valid and reliable, but economical way. This is done with the questionnaire fsmpK in which the mentioned competence classes (social, cognitive, specialized, personal) are operationalized. Each competence class consists of several ability concepts (e.g. the competence class “social competence” includes ability concepts like “autonomy”, “ability to communicate“, etc.). With respect to these ability concepts, the apprentices rate statements (6 statements for each ability concept). Questions like “How often do you do this within your daily job-work?” have to be rated with a five point scale (very often, …, very rarely).
At the time of the proposal, data from about 1000 apprentices and two measurement waves (t1 = 2005, t2 = 2006) is available.
Findings:
First of all, we present the psychometric characteristics of our questionnaire based on the first measurement wave (t1, N=1096). Cronbach’s alpha is used for estimating the reliability. The respective values range between .63 and .77 (socc), between .69 and .78 (cogc), between .59 and .82 (perc) and between .75 and .92 (spec).
With linear structural equation modeling we show how the four classes of competencies interact empirically. Empirical tests of the theoretical model lead to different results, one of them is this:
appendix 1, figure 2:
This theoretically based model is empirically not confirmed as the fits are not satisfying and the path from spec to socc is neither high nor significant. Modification indices suggest to include a path from cogc to perc which strengthens the role of cognitive competencies (in contrast to the mentioned theory in which specialized competencies play the most important role). The data suggest that cognitive competencies play an important role in the apprenticeship of polymechanics. If this is the case, one could hypothesize that the competence class cogc plays an even more important role and include a new path from cogc to spec by removing the correlation.
The respective model, as shown in the next figure, has a much better fit.
appendix 2, figure 3:
At the time of the proposal, data from only two measurement waves were available. At the time of presentation, a third measurement wave will allow us to test interesting longitudinal hypotheses. These new results as well as models with latent variables will also be presented and discussed.
Conclusions, significance for the field:
With the presented instrument it is possible to measure competencies in a valid, reliable and economical way. The resulting models show how competence within the field of polymechanics is structured and how the four competence classes (social, cognitive, specialized, personal) interact with one another. As a particularly relevant result, we retain that the models show the high importance of cognitive competencies in the field of polymechanics. This finally leads us to discuss the requirements for future VET in this field.
|
