Bimodality of a factor

Hi people,

I am planning a project looking at socially desirable response types and have a sneaking suspicion, based on previous work, that I am going to get items loading onto the same factor both negatively and positively.

I may have to justify this as not being a statistical artifact of item wording, but of the effect of item content.

I was wondering if anyone knows of a good explanation of bimodality of a factor in PFA and where I can find it?

Last edited: