Now what?
Eric Schulz. June 2022. You've collected your data, now what?
This is the mark down and data sets for lecture 9.
- lindata.csv: CSV file of the structured bandits data.
- dynamicdata.csv: CSV file for Experiment 4 in our paper (Schulz, Franlin & Gershman, 2020)
- analysis.html: markdown html-file showing some simple analysis for the first data set.
In Epxeriment 4 of our paper, one group of participants got 10 rounds of structure (pos/neg), 10 rounds of random (ran) and then 10 rounds of structure (pos/neg). Another group (between-subjects design) got 10 rounds of random, 10 rounds of structures, and then 10 round of random again.
Our question was whether or not it mattered for participants to experience the structured rounds (pos and neg) early on. Adapting some of the code from the markdown, it is your task to perform some analysis to see if this is indeed true. You could, for example 1. check the data and manipulation, 2. check for and possibly exclude outliers, 3. produce some simple behavioral plots, 4. produce about 2 theory-driven plots, and 5. run 2-3 regression models to check for differences between the groups. What do you conclude?
It'd be nice if you could submit either just the code you have used or a markdown file showing the code and output etc. I don't mind if you're using R or Python for this. Markdown is easier for me because I can directly see the plots and outputs. Good luck!