Moving developmental research online: comparing in-lab and web-based studies of model-based reinforcement learning
Tasks, data, and analysis scripts for Nussenbaum, K., Scheuplein, M., Phaneuf, C., Evans, M.D., & Hartley, C.A. (2020). Moving developmental research online: comparing in-lab and web-based studies of model-based reinforcement learning. Collabra: Psychology.
We collected data from 151 participants on two tasks: the two-step task, as described in Decker et al. (2016) and the Matrix Reasoning Item Bank (MaRs-IB) as described in Chierchia, Fuhrmann et al. (2019).
Versions of both tasks, coded in jsPsych, can be found in the "tasks" folder. Please note: the tasks were designed to be hosted on Pavlovia. As such, they will not run locally unless the Pavlovia-specific code is commented out.
1. Developmental two-step task
This sequential decision-making task was originally described in Decker et al. (2016), and is based off of an adult task originally described in Daw et al. (2011). Participants make a series of sequential decisions to try to gain as much reward as possible. In this version, on each trial, participants first must select a spaceship, which then transports them to one of two planets where they can ask an alien for space treasure.
2. Matrix Reasoning Item Bank (MaRs-IB)
This abstract reasoning task was developed by and originally described in Chierchia, Furhmann et al. (2019). All stimuli used in the task are from the OSF repository set up by the original study authors.
Participants complete a series of reasoning puzzles within an 8-minute time limit. The jsPsych version of the task uses one of the color-blind friendly stimulus sets as well as the "minimal" distractors described in the original manuscript. It was coded by the Hartley Lab for use online with children, adolescents, and adults.
Data and analysis code
All raw data and analysis code can be found in the "analysis_code_and_data" folder. All analyses and results reported in the manuscript can be reproduced by running the R scripts (for all data summary statistics and regression analyses) and matlab code (for the computational modeling of the two-step task data).
The output folder contains all model results and generated figures.
For questions, please contact email@example.com.