Question

Often it is impossible to use experimental research designs in Political Science, as many political variables would be impossible to randomize (ideology, regime, etc.). How can we then test casual hypotheses if we don’t assign treatment and control like we’d do in an experimental setting?

Answer 1

Quasi experiment-Quasi-experimental designs can be either a nonequivalent control group design or a before-and-after design. Nonequivalent control groups can be created through either individual matching of subjects or matching of group characteristics. In either case, these designs can allow us to establish the existence of an association and the time order of
effects, but they do not ensure that some unidentified extraneous variable did not cause what we think of as the effect of the independent variable. Before-and-after designs can involve one or more pretests and post-tests. Although multiple pretests and post-tests make it unlikely that another, extraneous influence caused the experimental effect, they do not guarantee it.

Quasi experimental study used to estimate the causal impact of an intervention on target population without random assignment. Quasi-experimental research shares similarities with the traditional experimental design or randomized controlled trial, but it specifically lacks the element of random assignment to treatment or control. Instead, quasi-experimental designs allow to control the assignment to the treatment condition, but using some criterion other than random assignment.

In quasi experiment uses two techniques difference in differences and instrumental variables.

*In a quasi-experiment you try to transform your data to accomplish the same effect--only the predictor and response are correlated. This has the potential to be as enlightening as a randomized experiment, but they require transformations that are difficult to identify, assumptions that are difficult to test, and give results that are difficult to interpret.

Structural equation modeling (SEM) family of statistical methods, suitable for determining causality  which uses SEM as a main statistical tool set and only secondary data.causal conclusions we can able to infer from the results of the analysis.

Here we use quasi experimental research to determine the causal relation in political science.