There is a transitionary nature within political science, which requires a continual growth of the literature we compound. Geddes focuses on the overarching research design. She includes how to ask and choose research questions, the development of theory from which you derive hypotheses, or testable implications with which you can gather data on, and the ways in which cases are selected which allow for those data to be drawn and logical inferences to be made.
She approached theory building in two ways, one being that there is attention paid in a very disciplined way to the way that we build upon already existing theories allowing for compound, complex theories to come to light in a methodologically sound way. The second, is the need to have overarching, universal standards and norms in research design as a whole. These universal standards and norms hold political scientists to standards that allow for transparency, reliability, validity, and ethical research to be added to the literature that allows us to understand the political paradigms surrounding us.
Moving from theory building, Geddes suggests that the cases we choose affect the answers we find, warning of two specific types of selection bias. The first being the selection of our dependent variables arguing that selecting on cases on the dependent variable is dangerous and leads to sloppy and cherry picked cases that do not always stand when held to the total population of cases involving out dependent variable. Overall, she believes that we can do this, but that we should be very cautious when doing so. Second, she warns of regression to the mean issues, arguing that when this tactic is used, we tend to see changes that did not really occur. Geddes warns that selecting cases should be done very carefully and methodically, that allow for logical implications to be made rather than running the risk of selection of cases that would lead to false conclusions and miscommunications of data results. Logic of inference must be followed and not undermined, or the research cannot stand and should not be accumulated amongst the growing literature.
On another note, Geddes spends a substantial amount of pages considering the amount of, and the importance of case study evidence in comparative politics. This is argued as a way to create and build upon a substantial and firm foundation in the field. While comparative is a method of discovering empirical relationships, it is not the scientific method, and it is not a method of measurement.
Morgan,
Good summary of what she’s saying. Can you add to you ‘concepts’ post by pulingl out the most important concepts that she’s using throughout the book? You can do this in bullet points — identify the concept, maybe highlight it in bold, quote her definition of this concept, and add your own comments about how she uses it throughout the book. For each concept, is it useful? Is it different from what we see elsewhere in the literature?
Hi, Morgan,
Thank you for this post, especially this section: “On another note, Geddes spends a substantial amount of pages considering the amount of, and the importance of case study evidence in comparative politics. This is argued as a way to create and build upon a substantial and firm foundation in the field. While comparative is a method of discovering empirical relationships, it is not the scientific method, and it is not a method of measurement.”
One of the trickier aspects of comparative politics may be the search for empirical relationships while using methods that are not necessarily measurements of something material or naturalistic. I like that you picked out the author’s contrast between the empirical relationships found in comparative political study work- or the empirical patterns- and the very different approach to “measurement” that political science has to other sciences.
Yes, I agree!
Thank you! I think that Geddes work in showing that the measures and the empirical work needed to be done in comparative is so applicable to all of political science as a whole. It is difficult to really conceptualize and meausre the concepts and relationships that we are dealing with. Because of this, I feel that not only does Geddes contribute to the methodological workings and structures inside of comparative politics, but that she also speaks broadly to the entirety of political science methodology.