Skip to content

Latest commit

 

History

History
65 lines (42 loc) · 3.35 KB

README.md

File metadata and controls

65 lines (42 loc) · 3.35 KB
Raw

"Self-Regulation Ontology" Dataset

Stroop Task

On each trial, subjects will be presented with a color word written in ink that is either congruent with the word, or incongruent with the word (e.g., blue written in blue =congruent, blue written in red = incongruent). Subjects are instructed to quickly and accurately respond via key press what the ink color of the word is. https://expfactory.org/experiments/stroop/

For each trial, we use the following features:

TODO: the following table needs to be updated (See StroopSRO data module)

SRO Feature Description CogPonder Name
worker_id the identifier of the participant. subject_ids
... the trial number. trial_ids
stim_color the ink color of the word (blue, green, red). stimuli[...,1]
stim_word the word (blue, green, red). This is the expected response. stimuli[...,1]
condition congruent or incongruent. trial_types
correct whether the response was incorrect=0 or correct=1. correct_responses
key_press response, which is expected to be the color of the word (blue=66, green=71, red=82). responses
rt response time in millis. response_steps

Adaptive N-Back

A variant of the n-back task, each trial consists of letters presented in successive order. Subjects must press a button if the current letter matches the letter that occurred n trials ago (regardless of capitalization). N varies according to a staircase tracking method that increases as subjects accurately respond and decreases as subjects make errors. https://expfactory.org/experiments/adaptive_n_back/

TODO: the following table needs to be updated (see NBackSRO data module)

The SRO-2back interface provides the following features from the Self-Regulation Ontology study:

  • input X_{ij}: previous N+1 symbols for the subject $i$ and trial $j$. The last symbol is the current trial. For each subject, $X_i$ is a 2-dimensional vector of integers of shape ($N_{\text{trials}}$, 3).
  • trial_type: Correct match, incorrect match, correct-non-match, incorrect-non-match for each trial $i$.
  • is_target: whether the trial $i$ is a match; it is a boolean.
  • output response: the response of the subject for the trial i; it is a boolean.
  • response_step: the response step of the subject for the trial i; Response step is an integer and represents response times in 50ms steps. This step duration is a hyperparameter of the data module.

License

From the original repository: https://github.com/poldrack/Self_Regulation_Ontology/

Shield: CC BY 4.0

This work is licensed under a Creative Commons Attribution 4.0 International License.

CC BY 4.0

This folder contains all of the data releases and derived data. Each folder is named with the sample name and the release date. The samples are:

  • Discovery: 200 subjects assigned to discovery set
  • Validation: 300 validation subjects
  • Full: Full sample of all complete subjects
  • Failed: non-completed subjects

These data should not be changed once being committed. Any additional derived data should be placed in the appropriate directory underneath the Derived_data folder.