Behavioural and Cognitive Task Dictionary

A dictionary of behavioural and cognitive tasks (Stroop, Implicit Association Task, n-back, Raven's Progressive Matrices), harvested from the Cognitive Atlas task ontology. The task counterpart of scales.

Usage

tasks

Format

A data frame with one row per task and 13 columns:

code

Slug used as the OSD scale_info$code: uppercase letters, digits and hyphens only.

name

Canonical task name.

acronym

Acronym from the Atlas alias field, when that field holds a real acronym rather than a synonym; empty otherwise.

atlas_id

Cognitive Atlas identifier (trm_* / tsk_*). Populated for every row, and the key that links a task to its Atlas record and to an Expfactory implementation, whose config.json carries the same id under cognitive_atlas_task_id.

description

Atlas definition text. Empty when the Atlas records the literal string "None".

citation

Description of one citation, not necessarily the canonical one: Atlas citations are unordered, and the Implicit Association Task's first citation is not Greenwald, McGhee & Schwartz (1998).

pmid

PubMed id for citation; empty when none.

url

Canonical Cognitive Atlas URL for the task.

n_conditions

Conditions recorded upstream (e.g. congruent / incongruent). Zero for most tasks.

n_contrasts

Contrasts recorded upstream. These are mostly names ("response time to incongruent vs. response time to congruent") rather than computable definitions.

n_computable

Contrasts carrying real condition weights, and so expressible as OSD scoring. Non-zero for 7 tasks in the whole ontology.

indicators

Comma-separated normalised indicators (response_time, accuracy, score, neural, rating). The upstream field is free text with 225 distinct values, so it is mapped to this small vocabulary and dropped when it does not map.

text_ok

TRUE when the name is safe to match against manuscript prose. Present for consistency with scale_meta; unlike PhenX entries, Atlas task names are instrument names rather than topic labels, so this is TRUE for every current row.

Source

Cognitive Atlas (https://www.cognitiveatlas.org), harvested from its public task API. The Atlas has no per-task licence field: its own content is CC-BY, but tasks it merely names (WAIS, Conners) are proprietary. Nothing here grants a right to administer a task. Rebuild with data-raw/tasks.R.

Details

A task is not a different kind of object from a scale. It is expressed in the same OSD format, with trials as items and conditions as dimensions; see inst/osd/tasks-in-osd.md for the mapping and inst/osd/STROOP.osd for a worked example. There is no separate task format.

Entries that OpenScales already defines are dropped, because the Cognitive Atlas catalogues anything a participant does and so lists questionnaires (CES-D, MMSE, Pittsburgh Sleep Quality Index) as "tasks". An OpenScales definition carries items, subscales, scoring weights and often a reliability estimate, where the corresponding Atlas record carries only prose, so the OpenScales record wins and the duplicate is removed. 23 of the 857 Atlas tasks are dropped this way.

The Atlas records no design parameters, so no parameters block is harvested. Those are authored by hand per task.

See also

scales for questionnaire instruments, scale_meta for their item-level detail.