Dependent and Independent Variables

The text below is a transcript of the video above, edited with AI assistance for clarity and flow. Where the two differ, the video is the authoritative version.

You’ll learn to

Dependent and independent variables are building blocks of causal research design. If you want to ask a question about causation, it is critical that you know how to identify the dependent variable and the independent variable in a research question and hypothesis. In this video you will learn to:

  • Distinguish the dependent and independent variables in a causal hypothesis.

Example 1: Health care systems and policy responses

We start with the example of health care systems and policy responses, and identify the dependent and independent variables in it.

Research question: How did countries’ health care systems affect their abilities to mitigate the impact of the Covid-19 pandemic?

A theory is a plausible answer to this question, a tentative speculation about the answer.

Theory: Countries with public health care systems took advantage of their centralised bureaucracies to implement more effective policy responses.

A hypothesis is a testable implication: a statement that describes what we should observe out in the world given the theory.

Hypothesis: Countries with public health care systems had lower Covid death rates than countries with private health care systems.

The unit of analysis here is the country. The “what” of the study is the country: we are comparing countries with and without public health care systems, and we are comparing the Covid death rates of those countries.

Countries differ from each other in a lot of characteristics. Population is one such characteristic. GDP per capita is another, and the land mass of a country is another. Countries vary widely in these characteristics, which is why in empirical research we call them variables.

In this study about the effect of health care systems on policy responses to the pandemic, two variables take centre place:

  • Health care system type (public v. private).
  • Covid death rate (a numeric quantity, data that can be collected from these countries).

The Covid death rate is the dependent variable. It is the outcome of the study: it is what is being changed by the independent variable. The health care system type is the independent variable: the variable driving the change.

Example 2: Economic conditions and electoral outcomes

Let’s now look at a second example, so that you fully understand the ideas of the unit of analysis and of the dependent and independent variables. This one is about economic conditions and electoral outcomes.

Research question: Do economic recessions increase support for populist parties?

Theory: Economic recessions increase resentment against democratic institutions and therefore make the rhetoric of populist parties more appealing.

A hypothesis, a testable implication, an observable pattern that follows logically from this theory, is the following.

Hypothesis: Elections held in recession years display higher vote shares for populist parties than elections in non-recession years.

Here the unit of analysis is the election. We are comparing elections held in recession years with elections held in non-recession years, and we are comparing the outcomes of these elections: how much support there is for populist parties in each election.

The outcome, the vote share for populist parties, is the dependent variable. The economic growth rate in the election year is the independent variable. The independent variable drives change in the dependent variable: the growth rate drives change in the vote share for populist parties, according to this hypothesis. Whether the hypothesis will be confirmed or not depends on the results of whatever study we design to answer this research question.

Recap

  • Variables are characteristics of the units in our study. The units of analysis have several characteristics in which they vary; these are the variables.
  • The dependent variable is the outcome of our study. It is affected by the independent variable.
  • The independent variable is what’s driving the change. It affects the dependent variable.

These are concepts that you need to understand fully, not only when you are formulating causal research but also when you are assessing, reading about or trying to critique it.

Why this matters

Dependent and independent variables are the building blocks of causal research. Everything that follows, the design, the statistical tests, the plots and charts that we make, all of these choices hinge on properly identifying the dependent and independent variables.