Domain-specific vs. domain-general collective efficacy

To develop collective efficacy scales, I believe we should first come up with some specific domains or contexts. As bandura claimed, domain general measures create serious probems such as predictive relevance. Thus, it should be context- and domain-specific, rather than measures assessing global collective confidence. It seems that we are not taking into account this issue. I am not sure what kind of collective efficacy scale we are trying to develop now (I know we have talked about collective efficacy in community settings and etc., but it is still not clear).