Running web-based experiments
Researchers are increasingly turning to web-based methods for collecting experimental data. The COVID19 pandemic has further accelerated this trend. The goal of this mini-course is to give participants an introduction to web-based experimentation. In the first part of the course we will discuss advantages and disadvantages of running experiments in-lab vs online, including a comparison of different data collection platforms like Amazon Mechanical Turk and Prolific Academic. We will also provide an overview of open science best practices, including transparent data sharing and pre-registration. The remainder of the course will be hands-on: using a running example from the experimental pragmatics domain, participants will learn to use git/GitHub for project management; modify an existing experiment; post the experiment to a data collection platform; and conduct basic data visualization/analyses using R.
Instructors: Judith Degen and Sebastian Schuster
Schedule
(all times are PST)
- 9:00-10:30: Introduction
- How do crowd-sourced experiments work?
- What kinds of experiments can be run online?
- Example study
- Using GitHub for research projects
- Preregistration and open science
10:30-10:40: Break
- 10:40-12:30: Tutorials
12:30-1:00: Lunch break
- 1:00-2:00: Tutorials (continued)
- Posting the experiment
- Testing the experiment
- Downloading and visualizing the data
- 2:00-2:15 Final discussion and Q&A