Yes, we need quantitative analysis
Honestly, I didn’t think much about the research methods that are popular in the discipline of political science when I began my preparation to return to graduate school. This might have something to do with the fact that I had only minored in the subject as an undergraduate, and therefore, had not encountered any of the heavier quantitative analysis that is common in the social sciences (e.g. sociology, economics, and psychology).
Once admitted to my program the daunting reality set in during the first semester that I needed to complete a minimum of two statistics courses for the degree. While I have always performed well in math, I knew that statistics is more than math and I felt anxious about the classes because I didn’t know what to expect. Thankfully, the combination of a quality instructor and comprehensible books translated into success. I completed the assignments and grasped many of the concepts (e.g. chi square, t-test, linear regression) in my effort to achieve a functioning level of literacy. Having to do some of the computations by hand really helped me when I think back.
Since my dissertation employs mixed methods and the survey I administered has a number of different questions, I had to learn about even more concepts (e.g. factor analysis) on my own so that I can analyze the data I collected. This has been challenging for me, but it’s an experience I’ve grown to appreciate – especially those moments when concepts click in my head. As my understanding of statistics continues to grow, not only does the methodological approach become less intimidating, but I also see the important role it plays in the advancement of knowledge.
Many people scoff at statistics because they think all numbers are subject to manipulation in furtherance of an agenda. When a trained political scientist tests a hypothesis by measuring the association of two or more variables, he or she is supposed to be executing a research design that is value free. Furthermore, his or her work will eventually be assessed by peers who study the same phenomenon. The point is that while I pick the variables I want to examine and measure the relationship between them, I can’t (nor do I want to) dictate the end results. This is what credible academics do – they report on, not manipulate, their findings.
