In the previous reaction post, I introduced the important new book A Model Discipline by Clarke and Di Primo. There I concluded with a note about how a multi-method approach provides a compelling reason for the combination of inductive and deductive analytic strategies. In this post I will discuss in more detail their argument for the varying purposes of models.
Clarke and Di Primo adopt a remarkably pragmatist approach to evaluating research. I am certainly sympathetic to this approach to epistemology and research methodology. I am pragmatist, through and through, with my research in public management. My reaction to most research in political science is "what pragmatic difference does this make?" It is worth considering the variety of purposes available for political science research, though.
The four purposes of theoretical models they propose are: foundational, organizational, exploratory, and predictive. This contrasts with the four purposes of empirical models: theory testing, prediction, measurement and characterization.
I won't get into the specifics of each purpose. This seems like a fine list. There is always the question of whether the list of exhaustive (why not an aesthetic purpose for theoretical models?) -- but that is largely beside the point. I fully support the call to pay attention (and state clearly in papers) the purpose to which a model is put. I see a place for most of these purposes. I think it largely misses the point of the AJPS decision to require empirical tests along with theoretical models, though.
I suspect a number of the editorial board that made the requirement that theoretical models be accompanied by empirical tests would largely agree with this list of purposes. The disagreement is over the match between the purposes and the publication outlet. I remember the reviews I got back from AJPS (before this policy, on a paper with a limited theoretical model -- and that may be kind -- and an empirical test). One reviewer gave it a weak review because he or she thought the journal should only publish the summative article of a line of research -- and this was clearly an early step. Another reviewer gave it a strong review because it was a novel, early step in a new research agenda. The reviewers simply had different views of the purposes of the AJPS and the sort of articles that should be included. To say there are different purposes does not imply that all purposes are appropriate for all outlets.
I think it would be good to have a debate over the appropriate purposes of publications in AJPS and other top tier journals. In public health, for example, their top journal (AJPH) regularly publishes literature reviews. There is NO chance that such an article would make it into AJPS. Why is this the case? This is not the fight Clarke and Di Primo want to start (more on that in the final review post) but it is the sort of question their pragmatic perspective raises. The argument they need to make is that the purposes of theoretical models (or some of them) match the purposes of the AJPS. This is different than just saying that there are different purposes.
I strongly suspect that my pragmatic orientation to evaluating research would be quite different from theirs. I would like to see more organizational and measurement work, for example. I have a high threshold for foundational and exploratory work -- which I think is better suited for field journals and white papers.
In the next post, I will discuss the implications of this pragmatic approach to the reviewing process.
No comments:
Post a Comment