Assistant Professor | Bloch School of Management

Plan of the day

  • Homework questions
  • Designing (real world) experiments

The basics of A|B Testing—also called the randomized controlled experiment.

Start with a simple question, and a narrowly defined outcome.

Make the manipulation of \(x\) clear, unambiguous, and meaningful.

Choose a representative (generalizable) sample.

Randomly assign participants to the control and to the treatment.

Have multiple people analyze the same set of data.

Pre-register your research design.

Lets do an experiment!

A couple of things about experiments…

Between vs. Within Designs

A couple of things about experiments…

Sampling strategy matters

A couple of things about experiments…

You really never get true random assignment, why?



What's next?

  • Next Class: Limited dependent variable models
  • Deliverables: Homework 8 Due 9:00AM Thursday