1

PS 300 Political Analysis

The Scientific Method Scientific inquiry involves both theory (logic) and observation–we are interested in an understanding of why things occur (theory) and some evidence that they indeed occur this way (observation). In the process of doing social science research, we are constantly moving between the theoretical and empirical (observational) levels. We can present this process in the following diagram:

The traditional model in social science starts at the theoretical level. Hypotheses (which are statements about the relationships between events) are derived from a theory, then subjected to empirical testing. This process is called deduction. An alternative starting point is the level of observation. Researchers make observations, look for patterns in the data, then from these patterns come to some tentative findings. These findings are called empirical generalizations and are used to construct theoretical statements. This process is known as induction. The important point to remember is that in our search for knowledge, we will continue around and around the circle above. Social scientific inquiry involves both deduction and induction. The Classical Approach or the Traditional Model (Deduction) The traditional model of science starts as a deductive process, i.e., at the theoretical level. Theories are statements of invariant relationships between or among concepts. They are explanations of some aspect of our world, e.g., juvenile delinquency, socialization, or social movements. In order to assess the adequacy of our theories, we need to derive specific hypotheses from them to subject to empirical testing. Hypotheses are propositions in testable form. They are statements of relationship between two or more variables.

2

Before we can test our theoretical hypotheses (statements about concepts), we must specify the meaning of the concepts to be studied. Concepts can have different dimensions just as words can have different meanings. Say for example, that we are interested in studying political participation. What do we mean by “political participation?” I can think of several dimensions. One dimension entails conventional participation such as voting in an election. Another entails unconventional participation such as street demonstrations, boycotts, etc. The importance of thinking about the dimensions of concepts comes when you place these dimensions individually in your original hypothesis. You may find that what you think will cause participation in “conventional participation” will not have an impact upon “unconventional participation.” The process of conceptualization, then, involves: (1) defining what your concepts mean (are there several dimensions?), and (2) determining how your theory would work for these dimensions. After stating your hypotheses and going through the process of conceptualization, you next need to decide how you will measure your variables. This is the process known as operationalization. Inductive Approach (Grounded Theory) Whereas the deductive model starts at the theoretical level, an inductive approach begins with observations. Maybe a theory does not exist to adequately explain a particular phenomenon. Say for example that there was a dramatic increase in teenage drug abuse. To try and understand why drug abuse may vary over time, you could engage in a research project studying teens over a particular period. You would look for patterns, which affect drug use, e.g., unemployment, family circumstances, etc. When patterns are found, you make tentative conclusions about the factors affecting drug use and abuse. In this approach, we are trying to construct theories and develop concepts to explain something. Again, remember the circular process of the scientific method. After you construct a theoretical statement, it should then be stated in a hypothesis and tested empirically. After the empirical test you may need to make new generalizations and modify your theory. This process continues until you have found a theory, which indeed is made-up of statements that are invariant, i.e., laws.