Even though the number of investigations into effective pre-service teacher education at university increased in the last years, scientific knowledge on this subject is still limited (Darling-Hammond, 2016). This also applies to physical education (PE) teachers, where evidence is missing with regard to experimental analysis. Since teacher education is increasingly geared towards linking theory and practice, there is considerable inducement to conduct such research (Borko, Liston, & Whitcomb, 2006). Reflecting examples of real-life classroom practice (i.e. teaching examples) is one approach to this end (Seidel, Blomberg, & Renkl, 2013). From a general perspective, in accordance with Bandura’s (1986) social learning theory, reflecting examples supports learning new skills. Similarly, following theories of analogical reasoning, examples can contain an convey prototypical rules and approaches that help individuals to handle new complex situations (Gentner, Loewenstein, & Thompson, 2003). Renkl (2011) states that the success of reflection on examples is due to a better understanding of general principles and a better comprehension of the relevance of the principles of a learned concept. Even though these theoretical perspectives do not focus on teacher education in particular, we assume teaching examples can illustrate specific teaching aspects to be learned. Beyond, one core argument for reflecting on PE teaching examples refers to the biography bias: Pre-service PE teachers often made numerous experiences in physical activities and sports before beginning their teacher education. These experiences foster specific teaching beliefs towards physical activities and sports (Volkmann, 2008). These often compete against the theoretical and empirical grounding of PE teaching (Volkmann, 2008). Teaching examples that provide new insights into PE learning might challenge these biographical experiences.

With regard to experimental studies that investigate effects of examples in learning processes, there is several empirical evidence in well-structured domains such as mathematics (Renkl, 2011), but a lack of findings in less-structured domains such as teaching. Beyond, it is unclear what quality teaching examples should provide: Some approaches employ “good” teaching examples, others provide pre-service teachers with “problematic” teaching examples. In well-structured domains the impact of incorrect and correct examples has been empirically addressed. In particular, problematic examples are assumed to demand effortful cognitive activities (VanLehn, 1999). In less-structured domains such as teaching, however, the question whether good or problematic examples lead to different learning results is still open.

The aim of our study is to explore if reflection on teaching examples enhances PE pre-service teachers’ instruction planning. To this end, we use a measure that is closer to professional practice than knowledge tests and indicates the extent to which teachers can apply their knowledge in realistic classroom settings (Grossman & McDonald, 2008). In a first experiment, we investigate whether undergraduate PE pre-service teachers’ instruction planning and their teaching beliefs changes after they reflected on good teaching examples, on problematic teaching examples, or compared both types of examples. In a second experiment, we extend our findings and investigate whether comparisons of pairs of good teaching examples or comparisons of good and problematic teaching examples better foster the pre-service teachers’ instruction planning. The central topic of our learning unit is implementing learning environments that enhance students’ self-controlled learning. The implementation of self-controlled learning is of particular importance within PE, since being able to act in a self-controlled way within PE class and when exercising outside of school is a core learning goal of PE (Scheid & Prohl, 2017). Consequently, teachers should know how to provide a learning environment that enhances students’ self-controlled learning. Accordingly, we used features derived from Bund and Wiemeyer (2004) to develop both good and problematic teaching examples and to evaluate the pre-service teachers’ instruction planning.

Both experiments had a pre-intervention-post-design. In the first experiment we compared three learning environments to which the N=83 students were randomly assigned to: In the good teaching condition, students reflected on eight good teaching examples, in the problematic teaching condition, students reflected on eight problematic teaching examples and in the compare condition, students compared eight pairs of good and problematic teaching examples. In the second experiment the N=81 students either reflected three pairs of only good teaching examples or three pairs of problematic and good teaching examples. The learning phases followed a standardized introduction on self-controlled learning in PE and was stable across the conditions. In both experiments, the pre- and posttest of instruction planning assessed the degree to which the pre-service teachers’ reflections took features of self-controlled learning into account. The items of pre- and posttest were drafted in a parallel way. For example, while pre-service teachers had to devise instructions to learn diablo tricks in the pretest, the posttest demanded instructions for devil stick tricks. For all parallel items there was a corresponding teaching example in the intervention. To evaluate the students’ responses, a coding system was developed based on the features that enhance students’ self-controlled learning in PE (Bund & Wiemeyer, 2004). As the coding procedure was highly inferential, 30% of the pre-service teachers’ responses were double-coded independently by two raters. The agreement between these coders was substantial for both experiments (.78

In both experiments we found a significant main effect of time on instruction planning (experiment 1: F(1,79)=8.85, p=.004, r=.32; experiment 2: F(1,79)=118.45, p<.001, r=.60). Although pre-service teachers’ ability to plan instruction does not automatically explain their teaching behavior in real classrooms, we assume, at least for novice stages, mentally simulating real-life classroom instruction to be an important prerequisite for connecting theory and practice (Borko, 2004). In experiment 1 a significant interaction of time×condition (F(2,79)=4.90, p=.010) could be lead back to a higher learning gain in the compare condition compared to both the good teaching condition and the problematic teaching condition. Between the good and the problematic teaching condition, there was no significant difference. One explanation is that it is better for novice learners to compare examples in order to better abstract the learned concept. Even though educational research assumes problematic examples to provide a specific learning potential (VanLehn, 1999), in experiment 2, students who compared only good teaching examples outperformed their counterparts who compared both good and problematic teaching examples (significant interaction of time×condition: F(1,79)=8.04, p=.007, r=.43). Again, this might be due to the students’ prior knowledge: As their pretest results were rather low, explaining only good teaching might be less challenging than comparing a problematic and a good teaching example. Finally, we can support the assumption that pre-service teachers’ experiences and reflections – realized by reflection on examples – influence their beliefs (Richardson, 1996): In experiment 1 results indicated a significant main effect of time on both constructivist view (F(1,80)=5.80, p<.001, r=.41) and transmission view (F(1,80)=60.18, p < .001, r = .66). In the compare conditions the constructivist view increased and the transmission view decreased more than in the two other two conditions. In the presentation, we will further discuss the results with regard to implications on (physical education) teacher education.

Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Borko, H., Liston, D., & Whitcomb, J. A. (2006). A conversation of many voices: Critiques and visions of teacher education. Journal of Teacher Education, 57, 199–204. Bund, A., & Wiemeyer, J. (2004). Self-controlled learning of a complex motor skill: Effects of the learners’ preferences on performance and self-efficacy. Journal of Human Movement Studies, 47, 215–236. Darling-Hammond, L. (2016). Research on teaching and teacher education and its influences on policy and practice. Educational Researcher, 45, 83–91. Gentner, D., Loewenstein, J., & Thompson, L. (2003). Learning and transfer: A general role for analogical encoding. Journal of Educational Psychology, 95, 393–405. Grossman, P., & McDonald, M. (2008). Back to the future: Directions for research in teaching and teacher education. American Educational Research Journal, 45, 184–205. Renkl, A. (2011). Instruction based on examples. In R. E. Mayer & P. A. Alexander (Hrsg.), Handbook of research on learning and instruction (S. 272–295). London: Routledge. Scheid, V., & Prohl, R. (2017). Sportdidaktik (2. Aufl.). [Didactics of Physical Education] Wiebelsheim: Limpert. Seidel, T., Blomberg, G., & Renkl, A. (2013). Instructional strategies for using video in teacher education. Teaching and Teacher Education, 34, 56–65. VanLehn, K. (1999). Rule-learning events in the acquisition of a complex skill: An evaluation of CASCADE. The Journal of the Learning Sciences, 8, 71–125. Volkmann, V. (2008). Biographisches Wissen von Lehrerinnen und Lehrern. Der Einfluss lebensgeschichtlicher Erfahrungen auf berufliches Handeln und Deuten im Fach Sport [Biographical knowledge of teachers. The effect of biographical experiences on professional behaviour and interpretations in physical education]. Wiesbaden: VS-Verlag.