Multivariate Data Analysis
An important part of the job of psychological research is discovering and explaining causal relationships. Randomised experiments form the best basis for causal inference. However, experimental manipulation of the independent variable cannot be realised for many research questions for practical, ethical, or principle reasons. Lack of generalisability of the results is another constraint of lab experiments. This is why causal relationships are sometimes studied exclusively, preferably or additionally in field settings. Research in field settings varies in the amount of control that the researcher has over the study characteristics. There are true experiments in natural settings such as hospitals, schools, and factories. Other (observational) studies have a quasi-experimental design, without control regarding the assignment of research units to a treatment or control condition. In evaluation research the intervention or programme is often decided on and delivered by an external organisation, and in case-control-studies the researcher can only try to arrange an optimally matched control group for the cases.
Students will be able to:
judge conclusions from field experiments on all relevant aspects of research validity;
propose an adequate research design, given the research question or hypothesis, to prevent of remove threats to research validity; and
apply adequate data analyses for estimating treatment effects in field experiments, e.g. correction with propensity scores.
For the timetables of your lectures, workgroups, and exams, select your study programme.
Students need to register for lectures, workgroups and exams.
Instructions for registration in courses for the 2nd and 3rd year
For information on registration periods consult the bachelor course registration
Elective students have to enroll for each course separately. For admission requirements contact your study advisor.
For admission requirements, please contact your exchange coordinator.
Students are not automatically enrolled for an examination. They can register via uSis from 100 to 10 calendar days before the date; students who are not registered will not be permitted to take the examination.
Registering for exams
Mode of instruction
8 2-hour lectures and 8 2-hour compulsory work group sessions.
The work group sessions consist of case discussions, student presentations of validity threats in published empirical field studies, and data analysis for estimating treatment effects.
A written examination assesses students’ theoretical and factual knowledge of causal inference in field experiments. This examination consists of multiple-choice and essay questions (weight: 60% of the final grade). A short written paper assesses the application of this knowledge to the validity of the conclusions of a published field study (weight: 40% of the final grade). Adequate active participation during the workgroups is an overall requirement for being awarded a final grade.
The Faculty of Social and Behavioural Sciences has instituted that instructors use a software programme for the systematic detection of plagiarism in students’ written work. In case of fraud disciplinary actions will be taken. Please see the information concerning fraud.
Shadish, Cook & Campbell (2002). Experimental and quasi-experimental designs for generalized causal inference. Boston: Houghton Mifflin.
A number of additional papers on the estimation of treatment effects in field studies according to the Approach of Donald Rubin.
Dr. Peter de Heus