The continuous random dot-motion (CRDM) paradigm

This repository contains the task code for running the continuous random dot motion (CRDM) paradigm, with an optional auditory paradigm in the background. Results from the CRDM task will be first presented in:

Ruesseler M, Weber LA, Marshall TR, O'Reilly JX, Hunt LT (bioRxiv Preprint, 2022). Decision-making in dynamic, continuously evolving environments: quantifying the flexibility of human choice.

Summary

In the continuous random dot motion (CRDM) task, participants are presented with a continuous (several minutes long) stream of moving dots. Their task is to press a button whenever they think they are currently in a period with average motion coherence to the left or to the right. The task code saves their responses (hits = correct detection of average coherent motion; error = detection of average coherent motion, but to the wrong side (right vs left); miss = no button press during average coherent motion; false alarm = button press during average incoherent motion), reaction times, and the points and money they win. Participants receive feedback for every response (the fixation dot changes colour according to the four different response types).

The repository contains all necessary functions, but needs Psychtoolbox (PTB-3) to be installed. The code was developed using MATLAB R2019b and tested in the OHBA EEG lab in Oxford.

Getting started

0. Clone this repository and add it to your Matlab path (including subfolders).
1. Modify the function crdm_set_task_options to change general task parameters - including any local settings such as keyboard indices.
2. Modify the function crdm_define_conditions to implement your own experimental design.
3. Set the variable basePath in crdm_run_experiment to where you want to save stimuli and participants' behaviour.
4. Run crdm_run_experiment.

Main functions

The main script performs all steps necessary to run an experimental session of the continuous dot-motion task:

crdm_run_experiment

The script will send prompts to the Matlab command window where you can enter the ID of the participant, session number, whether you want to record EEG data (and thus send triggers) etc. It will automatically save all stimuli and participants' behaviour for this session in a separate folder for each participant ID.

For every experimental session, you first need to generate the visual (and auditory, if desired) stimuli. The function crdm_create_session creates all stimuli based on the settings in the task options and your experimental design. Note that currently, you need to add your predefined auditory sequences to each block after creating the session, but you could easily add a function that creates the tone sequences automatically (depending on your auditory design).

Important: The stimuli should be generated on the computer on which the task will run (i.e., in the lab where you will test participants). This is because some stimulus parameters depend on local screen parameters.

However, you don't need to generate stimuli just before running the experiment - you can also do this ahead of time and only "play" the session later. Generated sessions are automatically saved in the participant's folder and can be loaded later to run the task.

The main function that runs the task is crdm_play_session. It takes a session and the general task options as inputs, as well as a base path pointing to where interim results should be saved (in case a session doesn't run through). The function implements several features that can be switched on and off using flags, e.g. options.doTones (whether to play an auditory sequence or not), options.doTraining (whether we want to signal response periods to participants by changing the colour of the fixation dot during training sessions), etc. All these flags are defined in crdm_set_task_options, except doEEG (which you will be prompted for) and doDebug (which determines whether the task will be run in full screen or in a small window for debugging). Note that the visual display only works properly if you created the stimuli in the same debugMode as you run the task later.

Data saved

The code saves the variables session and options with a session- and participant-specific name in the participant's folder, ending in "stimuli.mat", whenever you create a new session. After running an experimental session, it saves the variables respMat (which contains all responses and reaction times of the participant) and S (which contains all relevant session information including timings and trigger values) as a mat file ending in "results.mat".

Additionally, the code saves respMat and S after every block within a session (with a block-specific names, also ending in "results.mat") so that single blocks can be analysed in case the session cannot be completed successfully.

Auditory paradigm

The current task code will loop over a sequence of tones (defined in terms of their pitch) per block and for every tone, it will also create a silence which serves as the ITI. The parameters of the tones and ITI (the durations) are defined in the main task options. All that the code needs is a vector with pitch values for every block that has to be saved in a subfield of the session struct:

session.blocks(iBlock).audSequence

This allows, for example, to play an auditory mismatch negativity paradigm with deviants defined by unexpected changes in pitch. If you want to implement a duration MMN instead, you need to change the section of the function crdm_play_session where the tones are created via the loop over the entries of session.blocks(iBlock).audSequence (e.g., you can loop over a vector with durations instead).

Requirements

You need MATLAB and Psychtoolbox to run this code. Everything else is included. The code was developed using MATLAB R2019b.

Contributors

Maria Ruesseler, Laurence Hunt: wrote the initial version and core functions

Layla Stahr: adapted the code to run an auditory paradigm in the background

Lilian Weber: refactored the code into its current form