Randomized controlled trials in the social sciences (PHD416)

The use of randomized experiments has become increasingly popular and prevalent in social science research. The US Department of Education has labeled randomized experiments as the "gold standard" in educational research. The World Bank often requires developing countries to use randomization in determining the assignment and use of new policy.


Course description for study year 2024-2025. Please note that changes may occur.

Facts

Course code

PHD416

Version

1

Credits (ECTS)

5

Semester tution start

Autumn

Number of semesters

1

Exam semester

Autumn

Language of instruction

English

Content

This course focuses on the methodology of randomization in social science research. We focus on questions surrounding the use of randomization. Why is randomization so compelling? What assumptions are inherent in randomized designs? What are the hidden challenges to randomization? Is randomization always the "best" empirical strategy? How does one design randomized experiments? Is clustering a problem to randomization?

Additionally the course address important implementation issues in randomized controlled trials in social sciences, such as attrition, fidelity and compliance. What can be done to assure low attrition, and high fidelity and compliance? The paper discusses the practical aspects of running a randomized controlled trial including a discussion of its growing importance in policy formation and evaluation.

Finally, during seminars the students will be introduced to mixed method designs for investigating why or why not an intervention is effective. The class discusses how mixed methods are necessary for understanding mechanisms, fidelity, and compliance. Particularly, qualitative data collection can inform the researcher about each of these issues prompting improved data collection and rigorous scaffolding of complementary research designs.

The ability to understand which interventions in social science is essential to achieve important sustainability goals such as reducing poverty, improving education, and gender equality. In addition, the course discusses recent research covering these topics.

The course will last for five days and combines different work methods. We use a mixture of intensive lectures, PC-lab exercises, and seminars. During lectures, students will engage in the learning objectives and discuss examples of sophisticated RCTs. During the labs, students will work with software to measure statistical power. During the seminars, RCT researchers will visit and present large-scale RCTs that they are working on, as practical examples. The students will get the opportunity to apply their new knowledge to their own projects and will present their research design. They will receive feedback for their presentations and will be evaluated based on a final term paper in which they apply the different methods discussed in class to either their own paper or analyze and existing paper with high relevance for their own research.

Learning outcome

Knowledge

Students should be conversant about randomized experiments and have the basic tools to plan and to conduct randomized experiments.

Skills

Students should be able to

  • Understand causal modeling
  • Understand relationship of randomization to causal modeling
  • Gain the statistical tools to analyze randomized experiments
  • Become aware of underlying assumptions, strengths, and weaknesses of experimental approaches
  • Become acquainted with key players in the development and implementation of randomized experiments
  • Gain the statistical tools to design randomized experiments
  • Understand other key issues in design and implementation of random experiments.

General competence

The student should be conversant in the world of causal modeling. They should be able to be consumers and producers of experimental research. They should have the background to develop and implement ethically sound and scientifically valid randomized controlled trials. They should also have proficiency in software designed to measure statistical power.

Required prerequisite knowledge

Participants have to be enrolled in a PhD program.

Recommended prerequisites

It is mandatory that the students have basic knowledge of quantitative methods at master level.

Exam

Form of assessment Weight Duration Marks Aid
Term paper 1/1 Passed / Not Passed

Term paper requirement, 8-10 pages, demonstrating knowledge and application skills.

Coursework requirements

  • Full participation during the whole course is required.
  • Generally active participation in discussions in general.
  • Term paper requirement

Course teacher(s)

Course teacher:

Eric Perry Bettinger

Method of work

The course will consist of five days of intensive lectures, PC-lab exercises and seminars. During lectures students will engage in the learning objectives and discuss examples of sophisticated RCTs. During the labs students will work with software to measure statistical power. During the seminars, RCT researchers will visit and present large scale RCTs that they are working on, as practical examples.

Students are expected to prepare for and review lecture materials on their own.

Open for

The course is open to interested PhD candidates at the University of Stavanger and other universities. Single Course Admission to PhD-Courses.

Course assessment

There must be an early dialogue between the course supervisor, the student union representative and the students. The purpose is feedback from the students for changes and adjustments in the course for the current semester.In addition, a digital subject evaluation must be carried out at least every three years. Its purpose is to gather the students experiences with the course.

Literature

The syllabus can be found in Leganto