Motivations for choosing teaching as a career: effects on general pedagogical knowledge during initial teacher education

Abstract

The authors first ask to what extent future teachers in Germany endorse teaching motivations indicated by the FIT-Choice scale. This includes reporting on the confirmatory factor analysis they carried out to examine and to replicate the FIT-Choice scale structure in the specific cultural context of Germany with a sample of 1287 preservice teachers. Secondly, they investigate the relationship between future teachers' teaching motivations and their general pedagogical knowledge (GPK) using a subsample of 130 preservice teachers whose GPK was tested twice. Here the authors examine effects of preservice teachers' motivations for choosing teaching on their GPK. Among other findings of our study (1) the FIT-Choice instrument's factor structure is replicated, (2) the motivation profile typical for preservice teachers in Germany is replicated, (3) evidence is provided that intrinsic motivation is positively correlated, whereas extrinsic motivation is negatively correlated, with GPK at the first occasion of measurement, (4) extrinsic motivation has a positive effect on learning gain, whereas intrinsic motivation has not.

Keywords:

• motivation
• preservice teachers
• teacher education
• general pedagogical knowledge

Notes

1. TEDS-M, carried out under the supervision of the International Association for the Evaluation of Educational Achievement (IEA), was a comparative study of teacher education and the first IEA study on tertiary education as well as the first international large-scale assessment of future teachers that included representative samples. The TEDS-M target population was mathematics teachers for elementary and middle schools in their final year of teacher education.

2. Since the test instrument will be used in future studies we do not include more than these item examples in this article. In case there is serious interest by other researchers doing research in the field of the topic under study we will be pleased to provide them specifically with a documentation of the instrument as it was applied in our study. This documentation (König & Blömeke, 2010) includes various materials such as the test item pool, coding rubrics, scoring instruction, empirically based information on item parameters, and test booklets. It allows other researchers to use the test independently from the authors who developed it. Please contact the initial author of this article for further information.

3. The authors who carried out DIDAKTUM have fulfilled the technical requirements necessary to demonstrate the use of ethical procedures in researching human participants.

4. In Germany, due to the tracked secondary school structure teachers are prepared for either the academically oriented Gymnasium (lower and upper secondary school) or the non-academically oriented tracks, i.e. general education-oriented Realschule (lower secondary school), the more basic skills, vocational-oriented Hauptschule (lower secondary school) or any combination of such non-academically oriented tracks in comprehensive schools (Gesamtschule, Sekundarschule, etc.), depending on the individual federal state; cf. König and Blömeke (in press). This variation is mirrored by the various teacher education programs offered at universities (see Table 1). In our study, future teachers pursuing a teaching career for the Gymnasium are assigned to ‘Secondary’, whereas future teachers pursuing a teaching career for non-academically oriented tracks are aggregated and denoted as ‘Lower Secondary’.

5. Following the findings of Watt et al. (2012), the motivational factor ‘job transferability’ was not used because of unacceptable Cronbach's alpha measures of internal consistency in the German sample (.563). In accordance with the interpretation of Watt et al. the job transferability would hold less meaning for German preservice teachers. For example the item ‘A teaching job will allow me to choose where I wish to live’ (B45) does not make sense in the German context because of different requirements specified by the federal states (cf. König & Blömeke, in press).

6. For this analysis, we did not model GPK as a multi-dimensional construct it was done by König et al. (2011) since it was not the main research focus of this article.

7. Also there are lower latent correlations between ‘enhance social equity’, with each of ‘make social contribution’ and ‘work with children/adolescents’, in our study compared with the smaller German sample from the Free University Berlin used by Watt et al. (2012). Therefore it seems the conjecture outlined by Watt et al. concerning translation differences should be reflected on. From our view, one reason for the lower correlations could be that teacher education varies across universities in Germany with regard to various aspects (e.g. entry selection), so it is plausible to have lower correlations in our study due to sample heterogeneity. The scales obviously do measure different aspects, which, however, comes more to the forefront when future teachers from various institutions are sampled as it was done in our study.

8. We did not compute latent correlations nor complex regression models due to small sample size (n = 130).

9. To be sure, we also computed partial correlations, i.e. correlations between the FIT-Choice scale and the difference score where we controlled for age, sex and GPA (Grade Point Average; in German Abiturnote). Results are added in Table 6, right column. Since these findings did not lead to a different interpretation, we do not discuss them in detail.

10. Another issue for future research will be to differentiate general pedagogical knowledge into subscales it was done by König et al. (2011) and to relate them to the FIT-Choice scale specifically, since a single general pedagogical knowledge score may mask relationships between FIT-Choice motivations and knowledge sub-dimensions. The knowledge test used here can be divided into content-related sub-dimensions as well as sub-dimensions of cognitive processes (see Figure 2). For instance, we assume that test items measuring procedural knowledge (see Figure 4 and 5) will be correlated more highly with motivations than test items measuring conceptual-analytic knowledge (see Figure 3).