I am interested to understand how students' perceptions of their school climate affect their sense of political and civic efficacy, which in turn impact on their prosocial gaming behaviors (see research model in attachment).

Firstly I decided to conduct a cluster analysis to classify participants into groups based on their perceptions of the openness of their school climate the op. It was identified 2 clusters (low and high), which were used as preditor (independent) variables.

The results are confuse. How should be interpreted the fact that low school climate -> prosocial gaming behaviors present a non-significant total and direct effect, and a significant direct affect (see results in attachment)?

I sincerely appreciate your help and insights!