Identify interesting Variation.

Political Science is about observing, describing and explaining variation, i.e. changes that occur in the political world. We seek to develop a systematical understanding of why, when, and where certain things happen and change. We seek to explain the variation in the political sphere of human existence and identify regularities.

Hence, the first step towards a research project is to identify interesting variation. You may be interested in a particular topic, state, region, or institution and should ask yourself: What kind of variation is catching your interest?

It is useful to distinguish two types of variation – temporal and spatial. Temporal variation refers to changes that occur over time. When we study temporal variation, we seek to explain why a certain situation, event or outcome occurred at a certain point in time, for instance: Why did a government lose its majority? Why did war break out after a long period of peace? Why did an international institution change its structure or doctrines? Why did a new actor appear on the stage of politics, after being non-existent until this point? Why did the discursive construction of a certain issue transform in popular debates? …

Spatial variation refers to variation across units. Units may be individuals, states, institutions, or other relevant entities. When we study spatial variation, we explain differences across the units of interest, for instance: Why do some individuals vote for right-populist parties, and others not? Why do we see a revolutionary movement in one state, but not in the other? Why does one international institution adopt new rules, whereas another one seems extremely reluctant to introduce reforms? …

Many papers study temporal and spatial variation at the same time. Others confine attention to one of the two. The table below gives an overview of research designs by type of variation. A research design without variation is impossible if our aim is to explain variation. Research designs studying only temporal variation in a single unit are called time series. Works that compare differences across units without a temporal component feature only spatial variation and are usually called cross-sectional. Research designs combining both are time-series cross-sectional.

Reading tips.

Understand your Variables

Once you have

1. Give your variables a name and understand the or dimension a . If you talk of war do you mean civil war. If you talk of a rise of populism. And what is populism, anyways. Search for related literature in relevant journals to find conceptualizations and definitions for the concept of interest.
2. Level of Measurement
3. Level of Analysis

Link your variation to other variables. Look for Covariation.

1. Y-based
2. X-based
3. Descriptive
   1. Decribe, Typologies
   2. Great examples are sartori’s typology of parties, or the typology of regimes in general, such as the

All empirical evidence of causal relationships is covariational in nature. A purported cause and effect must be found to covary. They must appear and disappear, wax and wane, or perform some other transformation in tandem or at some regular, more or less predictable, intervals.

Make a theoretical argument

Once Usually you draw on established arguments and theories, which you may test and/or refine. Theory means that

Covariation can be extremely deceptive.

If your work is primarily descriptive, you should think of the relevance of your descriptions

Draw your data

Think the world like a regression

You may conceive of the world. The standard regression equation is not only a statistical formula – It represents a mode of thinking. Y represents the outcome. It is caused by a combination of different factors. These different factors x1, x2, x3 … xN. Our aim as political scientists is to either identify causes, or determine the outcome. Finally, we have the Error Term E. E represents the variation that we cannot explain. Social relations are so complex that we as social scientists will never capture all relevant aspects, which always leaves with some cases that our x-list fails to explain the outcome. That said, if we do really well in dentifying the various influential causes, we At minimum, we can determine the ligeklihood that a given configuration will lead to civil war, regime breakdown, or turnover of government….

Generalize and select appropriate cases

1. Case Selection
2. How representative is your case. On the line or of the line?

Draw on Data

1. A key Quali of pol scientists is data analysis. Even in descriptive works

Make yourself familiar with the Literature

Follow the Standard Structure

Write Sections in an efficient Order

Grading Criteria

Resource Overview

Gerring, John.  2004.  „What Is a Case Study and What Is It Good For?“.  American Political Science Review 98: 341-54.

Kellstedt, Paul M, and Guy D Whitten.  2018.  The Fundamentals of Political Science Research.  New York: Cambridge University Press.