Scientific Method

To summarize up front: In this video, I discuss how the scientific method applies to politics.  In this class, we will learn one way to study politics called empirical research.  This type of scientific study involves asking questions based on what we observe in the world, coming up with a potential answer to that question based on our background research, and then developing a specific way to test whether that answer is wrong based on new observation.  This is not the only way to understand and study politics, but it is the focus of this class.

When we talk about science, we usually think of the hard sciences – biology or physics – and people running experiments and doing computer analysis.  It can be presented different ways, but the scientific method fundamentally comes down to a single process.  You have a question about the world, come up with a potential answer to that question, find a way to test whether your answer is wrong, do that test, then communicate your results so others can learn from your experience.  What’s important about this is that science is a process.  “Science,” doesn’t tell us what is right or true, but it does provide an organized way we can think about cause and effect and rule out explanations for what is going on in the world that don’t have much support.  Good scientific research helps us narrow down the number and type of questions we have about the world – and opens up new questions for us to ask.

So what does this have to do with politics?  In this class, we will look at politics through a scientific lens.  This doesn’t mean we will run experiments (although we could), but that we will apply this process, this way of thinking to political questions.  Just like “regular” scientists, political scientists start by asking questions about the world based on what we have seen, we work to define those questions clearly, place them in the context of what we already know (or think we know) about a topic, then think of a way to evaluate that question.

Let me focus in more on each part of that process.  When we talk about questions, our focus is on questions about cause and effect.  For example, How does the state of the economy impact the results of presidential elections?  What makes civil war more likely to break out in a country?  Under what conditions are UN peacekeeping missions more likely to succeed?  These questions are generalizable and probabilistic – we look at events that happened in the past and ask what they tell us about how likely certain outcomes are. 

After deciding the question we want to answer, we work to define what we mean in more detail in part by looking at how other political scientists have answered that question. This is one way political science (and social science more generally) can differ from other types of research – we need to explain what we mean by the terms we are using and justify how we are measuring them.  So what do I mean by the state of the economy – is it overall productivity or individual perceptions?  What is the difference between civil war and social unrest?  And how can you tell?  In this class, we’ll learn the best ways to find existing research and then categorize the different conclusions those scholars have drawn about definitions and causes in a literature review. 

We’ll use that information to develop a potential answer to our question – a hypothesis that we can test based on new research.  That’s all what we’ll do in the first month in the course.  Then for the second and third month, we’ll talk about different ways we can conduct those tests – we’ll spend a month learning about statistical analysis, but then we’ll also discuss running experiments, doing interviews, conducting research in historical archives, and more.  We’ll review about which types of research make the most sense for which types of questions so that you can produce an effective plan to test a question – what we call a research design.

Note that one thing all this means is that we won’t look at policy questions – what we think making one specific change to a law will do in the future – although these types of questions are definitely related to what we learn. (Understanding what generally works is a first step toward understanding what might work in a specific situation.)  We also aren’t going to look at moral questions (what is right or wrong).  These are of course incredibly important, but they require us to look at more than just what has happened to evaluate them.  Other types of questions and answers are equally valid, just different.

The last thing I want to note is that we aren’t going to go through the whole cycle of the scientific method in this class – we are going to write research designs, but not conduct and publish the analysis – because we don’t have enough time to do it all.  You will gain the skills you need to conduct this research, however, and I do hope that some of you will continue your research in other classes next semester.