Causal Evaluation Methods in Education Research

Make Evidence-Based Decisions
How do education vouchers affect learning outcomes? What is the impact of financial aid upon higher education enrollments? Does reducing class sizes really improve student performance? Is there a relationship between educational attainment and civic participation? How effective are school-wide reading programmes?
Course levelAdvanced Bachelor/Master/PhD
Session 130 June to 14 July 2018 
Recommended course combinationSession 2: Hands-on Anthropology and Ethnographic StorytellingWildlife Crime Analysis: Data-Driven Nature Protection
Session 3: Foundations of Strategy: How to Succeed as a StrategistOperations Research: A Mathematical Way to Optimize Your World
Co-ordinating lecturersIlja Cornelisz, Chris van Klaveren
Other lecturersTo be confirmed
Form(s) of tuitionLectures, lab sessions, group discussion
Form(s) of assessmentTake-home exam
ECTS3 credits
Contact hours45 hours
Tuition fee€1.000

Students and professionals with an interest in and aptitude for empirical research in the social sciences, in particular education policy. If you have doubts about your eligibility for the course, please let us know. Our courses are multi-disciplinary and therefore are open to students and professionals with a wide variety of backgrounds. 

Policymakers around the world are increasingly making evidence-based decisions about the use of scarce resources in such fields as education. Recent advances in research methodologies and data quality have vastly increased the potential for rigorous evaluation of the causal impact of interventions. Yet doing so often remains challenging methodologically, given issues of selection and attrition in social sciences research.

Several estimation methods have been developed in recent decades to help identify causal effects in the social sciences. Applying these techniques to educational settings, this course provides you with the knowledge and empirical skills you need to identify whether interventions are effective. You gain hands-on experience of evaluating randomized controlled trials, difference-in-differences settings, regression-discontinuity designs, instrumental variables approaches and panel-data studies.
 
As well as interactive introductory lectures, you take part in practical lab sessions using statistical software to replicate actual results from seminal papers. In structured group discussions, you critically evaluate these results and reflect on lessons learned. Assessment is by a homework assignment testing your competence, both conceptual and analytical, with each of the research designs covered.
The course is offered by VU-ACLA, a team of researchers specializing in causal evaluations of educational interventions.

At the end if this course, you:

  • Can distinguish between the various experimental and quasi-experimental methods in causal evaluation.
  • Understand the strengths and weaknesses of these methods.
  • Can empirically estimate and interpret experimental and quasi-experimental regression models using statistical software.
  • Can determine whether empirical studies provide valid evidence and so should be included in systematic literature reviews.

 prof pic 2

Chris van Klaveren is associate professor at the Faculty of Behavioural and Movement Sciences at Vrije Universiteit Amsterdam and co-founder/director of the Amsterdam Center for Learning Analytics (ACLA). He obtained his PhD in Economics and his current research focuses on adaptive learning algorithms, learning analytics and impact evaluation.

"Policymakers around the world are increasingly making evidence-based decisions about the use of scarce resources in such fields as education. Recent advances in research methodologies and data quality have vastly increased the potential for rigorous evaluation of the causal impact of interventions. Yet doing so often remains challenging methodologically, given issues of selection and attrition in social sciences research. In this course, students get hands-on experience with evaluating (quasi-)experimental interventions in education."

 

prof pic

Ilja Cornelisz is an assistant professor at the Faculty of Behavioural and Movement Sciences at Vrije Universiteit Amsterdam and co-founder of the Amsterdam Center for Learning Analytics (ACLA). He has a PhD in Education Economics and his current research focuses on learning analytics, personalized learning, school choice, and the interdependence of education and the labour market. His specializations are causal program evaluations, education economics and education policy.

 

A visit to the Dutch Inspectorate of Education (tbc), followed by a social event.

Murnane, R. J., & Willett, J. B. (2010), Methods Matter: Improving Causal Inference in Educational and Social Science Research, Oxford University Press.

As it is used throughout the course, you are strongly encouraged to purchase a hard-copy version of this book. Other necessary materials will be made available online.

A basic understanding of linear algebra and statistics. Since results from studies are replicated in lab sessions, some working knowledge of statistical software packages (SPSS and/or Stata) is also required.
Facebook-icoon   linkedin-icoon   Instagram-icoon