Participants for study 1 were 95 physical education teachers (53,1 % males, 38.11 + 10.96 years old) who filled in a questionnaire consisting of four vignettes. The first two vignettes described an autonomy-supportive and a controlling style (Reeve et al., 2014). After reading the vignettes, teachers rated their engagement in both styles (e.g., “This approach to teaching describes how I teach my students on a daily basis”, Reeve et al., 2014). In addition, teachers read two vignettes describing two different types of student motivation (autonomous motivation, controlled motivation) and filled out two items tapping into teachers’ belief in the motivation-dependent effectiveness [i.e., “An autonomy supportive (controlling) style works best for autonomously motivated students (controlled motivated students)”], and the absolute effectiveness of the teaching styles [i.e., “An autonomy supportive (controlling) style works best for all students”]. Given that many teachers belief that a teaching style will have differential effects among students with a different motivational profile, Study 2 examined whether this holds true such that the effects of autonomy-supportive and controlling teaching depend on students’ motivation. Three hundred and twenty students from 42 different classes out of two secondary schools participated in an experimental study (Study 2). The students were on average 17.28 years old (SD = 1.36) and 33.1 % (n = 106) were boys. Prior to the experiment, students completed a paper and pencil questionnaire on their motivation (BRPEQ; Aelterman et al., 2012). Students were then randomly assigned to an autonomy-supportive or controlling condition, in which they watched a series of five short film fragments (duration on average 1’30”) displaying situations that typically occur in a PE lesson (e.g., providing help or dealing with disruptive students) that either represented an autonomy supportive or a controlling teacher style. In between each video fragment students filled out a short questionnaire measuring perceived autonomy supportive (TASCQ; Skinner & Belmont, 1993) and controlling teaching (PCT, Soenens et al., 2012), need satisfaction and frustration (BPNSNF; Chen et al., in press), engagement (Skinner, Kindermann, and Furrer, 2009) and oppositional defiance (Van Petegem et al, 2013).

Study 1 revealed that teachers were more likely to believe in the motivation-dependent effectiveness of a teaching style than in its absolute effectiveness. Further, multiple regression analyses showed that teachers reported to be more autonomy support, if they believed in the effectiveness of an autonomy-supportive style for all students (β = .37, p < .001), rather then when they believed effectiveness was motivation-dependent (β = .17, ns). Similarly, the use of controlling style was predicted by the absolute belief in the effectiveness of a controlling style (β = .50, p < .001) but not by motivation-dependent belief in controlling style (β = -.08, ns). Study 2 revealed that students viewing an autonomy-supportive teacher reported more need satisfaction, less need frustration, more engagement, and less oppositional defiance compared to students viewing a controlling teacher. Main effects of student motivation showed that irrespective of whether students viewed the autonomy-supportive or controlling teacher, more autonomously motivated students reported higher levels of need satisfaction and engagement, and lower levels of need frustration. In contrast, more amotivated students reported less need satisfaction and more need frustration, more oppositional defiance and less engagement. Tests of the interaction effects between students’ motivation and teachers’ style showed that only four out of 24 were (marginally) significant and they were a matter of gradation. Students’ motivation affected the degree to which autonomy-supportive (relative to controlling) teaching yielded beneficial outcomes rather than completely altering (reversing) these effects. Together, the findings suggest that an autonomy-supportive style generally yields adaptive outcomes, even when motivational regulations are less optimal, and a controlling style leads to detrimental outcomes even if students are controlled motivated or amotivated. Hence, if teachers want to promote students’ motivation, they would do well to adopt an autonomy-supportive stance and avoid a controlling style, independent of students’ motivation.