For a high-level overview of data collection capabilities, see Collect Rich Data.
mindLAMP includes 30 cognitive games — brief, tap-based assessments of cognitive performance. Each is built on an established experimental paradigm from cognitive psychology and neuropsychology, and each records timestamped, structured output: not only whether a response was correct, but the reaction times and response patterns behind it.
The games are grouped below by cognitive domain; each links to its reference page for the full data schema, scoring, and configuration. For ActivitySpec identifiers and Cortex-feature status, see the Activity Reference; for the shared event structure, see Activity Data Format.
Attention & inhibitory control
| Game | What it measures | Primary output |
|---|
| Stroop | Selective attention and interference control | Stroop effect (incongruent − congruent RT) |
| Flanker | Attentional control and conflict resolution | Flanker effect (response-time cost) |
| Pop the Bubbles | Sustained attention and response inhibition (go/no-go) | Commission and omission error rates |
| Cats and Dogs | Attention, response control, and set-shifting | Accuracy and response time |
| D-Cog | Adaptive attention and response control | IRT ability estimate (0–10) |
Working memory
| Game | What it measures | Primary output |
|---|
| N-Back | Working-memory updating | d′ (sensitivity) and response criterion |
| Digit Span | Verbal working-memory span | Forward and backward span |
| Spatial Span | Visuospatial working-memory span | Maximum Corsi span |
| Memory Match | Short-term visuospatial recognition | Mean moves to clear; perfect rounds |
Processing speed
Executive function & planning
| Game | What it measures | Primary output |
|---|
| Tower of London | Planning and problem-solving | Problems solved in the minimum moves; excess moves |
| Wisconsin Card Sorting Test | Cognitive flexibility and set-shifting | Perseverative errors; categories completed |
| Water Sort | Planning and sequencing | Solved in the minimum pours; excess pours |
| Nonogram | Deductive reasoning and constraint satisfaction | Puzzles solved; errors |
| Trails B | Set-shifting (Trail Making B) | Completion time and accuracy |
| Jewels B | Set-shifting (Trail Making B analog) | Completion time and accuracy |
Learning & memory
| Game | What it measures | Primary output |
|---|
| Funny Memory | Associative learning and recall | Forced-choice recall accuracy |
| Memory Game | Sequence and spatial-location recall | Encoding and recall accuracy |
| Fragmented Letters | Visual perceptual identification | Audio responses (scored offline) |
Spatial cognition
| Game | What it measures | Primary output |
|---|
| Mental Rotation | Spatial visualization | Accuracy and reaction time by rotation angle |
| Sliding Puzzle | Spatial planning and problem-solving | Move efficiency (optimal ÷ actual) |
| Maze Game | Spatial navigation | Time and levels completed |
Language
| Game | What it measures | Primary output |
|---|
| Lexical Decision | Lexical access and word recognition | d′; word-frequency effect |
| Letter Logic | Vocabulary and hypothesis testing | Rounds solved; mean guesses |
Social cognition
| Game | What it measures | Primary output |
|---|
| Emotion Recognition | Facial-emotion identification | Correct identifications per trial |
Decision-making & risk
| Game | What it measures | Primary output |
|---|
| Balloon Risk | Risk-taking (BART) | Adjusted average pumps |
| Spin the Wheel | Reward learning under uncertainty | Risk-taking rate; final balance |
| Delay Discounting | Temporal impulsivity | Area under the curve; hyperbolic k |
Data & scoring
Every game produces an ActivityEvent with a session-level summary (static_data) and per-action detail (temporal_slices), so both the outcome and the process behind it are available in the raw data export. The metrics above are computed in-app and stored on each event (Fragmented Letters is the exception — it is scored offline from its audio responses).
A subset of the games — Jewels A and B, Spatial Span, Cats and Dogs, and Pop the Bubbles — also feed derived Cortex features for longitudinal analysis. The Activity Reference lists each game's ActivitySpec identifier and its current Cortex status.
Used in research
mindLAMP's cognitive games have been validated against gold-standard neuropsychological measures: the Jewels Trail Tests (digital Trail Making analogs) have been validated against the paper Trail Making Test in schizophrenia (Shvetz et al., 2021) and in major depression (Thérond, 2021). The broader smartphone battery has been evaluated in multi-site schizophrenia cohorts (Castillo et al., 2025).
These assessments have also been used in longitudinal digital-phenotyping research across schizophrenia (Liu et al., 2019), Parkinson's disease (Weizenbaum et al., 2021), and first-episode psychosis (Lakhtakia et al., 2022).
For the full list of studies, see Publications.