Impact: Causal Inference and Impact Assessment
Content
Much empirical research is motivated by questions of causation. Sometimes this question involves the relationship between passively observed variables in a data-set e.g., when analysing survey data, one may ask whether education ‘causes’ income, health or happiness? Sometimes the question involves explicit manipulation of conditions: do interventions such as treatments, stimuli or policies yield specific effects? In this form the question is very common in medicine, marketing and advertising, and all kinds of government intervention.
Whether data are generated by passive observation or by explicit manipulation, the question arises on what basis correlations can be interpreted as causal effects.
This clinic focuses on methodological issues involved in making causal inferences. It does so partly by comparing the strengths and weaknesses of different research designs e.g. experimental, quasi-experimental and non-experimental procedures. It also explores the practical and methodological concerns common to each design when assessing the impact or effects of interventions. These include taking into account known and unknown contaminating factors using randomization, statistical control and matching, maintaining sufficient statistical power, and identifying conditions that could undermine causal attributions, such as endogeneity.
This clinic will be supplemented with online learning materials which can be accessed after the event.
Prerequisites
Introductory understanding of statistics, e.g. the mean, standard deviation and standard error of a variable. If you are attending the clinic as part of our Researcher Development Initiative then you automatically fulfil the prerequisites.