I ran a MANOVA model, with 2 DV's (2 groups of adolescents) and 1 IV with 3 categories ("strategies"). I got a non significant result, with P. Value of 0.15 (for all categories). I know that if there is not an interaction effect one should stop at this point. However, the output (SPSS) also gave the 3 single ANOVA's, and one of them WAS significant with P Value of 0.05 (with a bonferroni correction) but with a low power of 0.4

What kind of conclusion can I give in such a situation ? Can I state the above or is it plain wrong/ insignificant to even include the information? I tried controlling for age to see if that would make a difference, but unfortunately it did not.

Can I say that overall there are no differences between the groups when comparing all 3 strategies used but there is a difference when comparing each strategy separately ? Given that I reduced Type error I when using bonferroni is there any room to state this in a publication?