1; p<0.01; ω2=0.19;), as well as a significant medium effect on total motivation (F(1, 111)=9.9; p<0.05; ω2=0.08;). No significant main effect of school type and of gender

neither on motivation in total nor on any of its subscales was found. The same holds for the rest of the covariates. Within subject effects ( Table 8b): both total motivation and all its subscales showed significant and strong interaction of their temporal development (or time course, TC) with group membership (CC: F(2, 222)=79.1; p<0.01; buy MDV3100 ω2=0.40; SC: F(2, 222)=59.2; p<0.01; ω2=0.34; IM: F(2, 222)=57.1; p<0.01; ω2=0.33; total: F(2, 222)=75.8; p<0.01; ω2=0.39). Significant, but only small up to medium main effects on motivation in total and on two of three subscales were found for the interaction “time course vs. school type” (CC: F(2, 222)=6.1; p<0.05; ω2=0.05; SC: F(2, 222)=11.2; p<0.01; ω2=0.09; total: F(2, 222)=8.1; p<0.05; ω2=0.06). Finally, significant small up to medium size effects were found for the threefold TC×GM×ST interaction as well as the fourfold ZD1839 chemical structure TC×GM×ST×GR interaction on each sub-dimension and motivation in total (TC×GM×ST; CC: F(2, 222)=12.7; p<0.01; ω2=0.09; SC: F(2, 222)=9.8; p<0.01; ω2=0.07; IM: F(2, 222)=3.6; p<0.05; ω2=0.04; total: F(2, 222)=5.9; p<0.05; ω2=0.05; for TC×GM×ST×GR

interaction: see Table 8b). Thus motivation developed differently in time for different treatment groups and school types. in particular, the strongest significant interaction 4��8C characterized by large effects on each subscales and motivation in total was found for the TC×GM interaction. As inspection of the time course (Fig. 2) clearly shows, the TC×GM interaction obviously is due to large differences between treatment and control group in development from before to after intervention (and not a possible

difference afterwards, i.e. from post to follow-up test). Thus, the results of the temporal development of motivation within subjects is consistent with the between subjects main effect of group membership on motivation. The influences of the threefold and the fourfold factor-interaction on each sub-dimension and motivation in total mean, that the positive motivation development of TG compared to CG was different for both school types ST1/2 considered (as clearly visible when comparing Fig. 3 and Fig. 4). This might be due to chance, but a plausible explanation is as follows: in school type 1 (“Realschule”) grade 10, where the instruction took part, is the last year in school; a general drop of motivation towards the very end of the schooling period is a well-known experience of many partitioners supported by the data ( Fig. 3, CG curve). This factor does not exist for school type 2 (“Gymnasium”, Fig. 4), hence the difference found. No influence of gender or of any interaction of gender and other factors neither on motivation in total nor on subscales could be discovered. The holds true for the influence of all other learner covariates considered.