[MUSIC] Hi, in this module, I'm going to talk about theory and evidence in social science. Social science theories seek o explain relationships between what we refer to as concepts. By concept we refer to a trait of a social entity that we just talked about. Concepts are abstractions, even though we may think of specific ones almost as if they were very concrete and real. Let's look at some examples. So if we have some social entities, it could be individuals or other entities. When we talk about concepts at the individual level, we might think of things like health Income and education. Now, we take these for granted but in fact they are abstract, so we call them concepts. At a broader level, we might talk about prosperity, transparency, openness as characteristics or traits of countries. So again, we distinguish between individuals on the one hand, and concepts that are relevant for them, and countries and larger scale units on the other. We'll be coming back to the very important distinction between scales in later lectures. Again, a theory to count as a theory should not only describe a relationship between two concepts. But it also explain that relationship. For example, we have a theory about the relationship between education and health. It should not only describe the relationship, and suggest that education influences health in some consistent way, but it should also suggest a mechanism, a pathway, by which education influences health. A statement about a relationship between two concepts that does not offer an explanation. Generally, we don't regard it as a theory, it's mainly a statement of an empirical irregularity or a law. Let's take a look at an example. So going back to this issue of the importance of theory offering and explanation for relationship. If we're thinking of the example of the relationship between education and health, somehow a theory must identify a pathway by which education is proposed to influence health. So there are many things by which education might influence health. And theories have been proposed. So we think that education influences behavior. Education also influences income. Now, income might also influence behavior. Income in turn influences health care and stress, and behavior influences stress, and then these somehow all combined to shake health. Specific theories might focus on a specific pathway going from education to health, and offering a account of why a particular pathway perhaps the one going from income through stress helps link education to health. So it's that linkage that explanation that helps make something a theory. To test our theories, we have to translate our concept into indicators. And then, from a hypothesis about the relationships between those indicators, that we expect to see out in the real world. By indicators, we refer to measures that can be collected in a systematic fashion in the course of a study. As opposed to the abstractions that they are based on, that is the concepts. So if we think about education and income and health, these things that we're interested in. If we were wanting to test a theory about their relationship, we would need to translate each of them into an indicator that we could go out and measure. So for example, education we might measure as numbers of years of schooling. There are other ways to measure education but for the purpose of the study, we must settle on one particular indicator or measure. So numbers of years of schooling. Income, there's lots of ways to measure income, might be typical in at least certain countries to measure income as annual learnings. In other settings people might ask, monthly earnings. And finally, health, we have to translate that very abstract concept health into something we can go out and measure. It's very common for people assessing health or trying to measure it to actually use self-rated health where people are simple asked, how healthy are you on a scale of one to five? If we collect our information, collect our data on the indicators, and then the relationship among the indicators is as predicted, or expected from the hypothesis, then we think of that as confirming our theory. Now, there's one issue I'd like to introduce, about different ways of developing theory, to help clarify the distinctions between disciplines. That is the distinction between Inductive approaches and deductive approaches in the development of theory. So in a inductive approach, theory generally starts in some ways from the outside. People observe the world around them, they look for patterns and regularities. This leads to generalizations which then leads to actual specifications of theory that we go out and try to test. A deductive approach starts with theory with carefully defined rules that are used to derive predictions and the these predictions are the basis for the design of experiments or possibly observational studies that try to test the original theory. So social science tends to be heavily, although not exclusively, inductive, where a lot of social science theory is developed as the result of people observing the world around them, noticing trends, and patterns that lead them to make generalizations, which then inspires the development of theory. If we go to the other extreme, we have disciplines like mathematics and philosophy which are almost entirely deductive, which don't even conduct experiments necessarily, they operate entirely within a world of theory. Then there are some part of the natural sciences and even some parts of the social sciences, that are also deductive, that start with strong theory. This is especially the case in economics, to some extent in political science. So these are the basic elements of theory in the social sciences and this helps lay the groundwork for our discussions in later lectures on how we design studies to test theories.
