Garf! Thinking too hard.
Sometimes after you learn more about a topic, you discover that something you said about the topic actually carried ramifications of which you were not aware. Or you could be thinking too hard and reading something into it that was not actually there. Now that I've introduced a borderline nonsensical abstraction, let me attempt to make it more concrete. I've been corresponding with a potential collaborator on a project that he has completed, but now needs to be analyzed. Therefore, I don't know anything about how this project was conducted, what the variables were, etc. But, apparently the project could be analyzed using SEM. I received two emails about the project that were relatively abstract and I attempted to concretize it (is that a word?) by thinking about how SEM could be utilized without knowing the exact variables. So I responded by speculating whether it would require parceling and whether there might not be more than one indicator per some of the latent traits (i.e, introducing elements of path analysis). What I did not realize until later was that parceling is a very controversial topic within SEM. To parcel or not to parcel? Apparently, that is a very loaded question. Furthermore, minimal numbers of indicators per latent traits is a no-no in SEM. So I think my innocent postulating could have insulted my potential collaborator. Hrm. Who knew that statistical techniques could be so "political?" Yesterday, he said he would write more on Friday, but he never did. Is the failure to write a function of my reply or simply reflective of a busy day with no time to write? See? This is where the second part of my earlier abstraction comes into play. Am I just imagining that an offence exists where no offence has been taken? I wish I knew the answer to that. I suppose I will have to wait and see.