Activity Suite Update: Enriched Data Output and Bug Fixes
mindLAMP's cognitive assessment suite has been updated with enriched data output, bug fixes, and improved mobile responsiveness. All changes to existing games are backward-compatible: existing data fields are preserved and continue to mean the same thing. New fields have been added alongside the old ones to support richer analysis.
What's Changed Now
The following changes are live in the current version of mindLAMP.
Enriched Data Output (Additive Only)
Several existing games now produce additional fields in static_data and temporal_slices. All previously existing fields are preserved and unchanged. If your analysis pipelines already work with these games, they will continue to work without modification. The new fields are available if you want to take advantage of richer data going forward.
| Game | New Fields | What You Can Do With Them | Documentation |
|---|---|---|---|
| Digit Span | forward_span, backward_span, total_raw_score, forward_trials_correct, backward_trials_correct | Compute standard WAIS-IV span scores directly. The forward-backward span difference indexes executive contribution to working memory. All legacy fields (score, correct_answers, bestForwardDigitSpan, etc.) are retained and computed the same way. | Digit Span docs |
| Spatial Span | max_span, total_score, mode, level_results, levels_completed, starting_span, total_levels | Standard Corsi block-tapping scoring. max_span is the primary visuospatial memory measure. level_results provides per-trial span, accuracy, and response time for error-pattern analysis. All legacy fields (correct_answers, score, StartTime, EndTime, etc.) are retained. | Spatial Span docs |
| Pop the Bubbles | levels array (with go_correct, go_total, go_missed, nogo_correct, nogo_total, nogo_wrong), score, correct_answers, wrong_answers | Compute commission error rate (nogo_wrong / nogo_total) for inhibitory control and omission error rate (go_missed / go_total) for sustained attention, the core clinical distinction in go/no-go tasks. Previously, go and no-go performance were lumped together. | Pop the Bubbles docs |
| Spin the Wheel | risk_taking_rate, net_earnings, mean_reaction_time, total_high_risk_choices, total_low_risk_choices, final_balance, starting_balance | risk_taking_rate is the primary risk preference measure (0.5 = no preference). Per-spin temporal_slices now include risk_level, win_amount, loss_amount, and balance_after, enabling analysis of post-outcome risk adjustment. Previously static_data was empty. | Spin the Wheel docs |
| Trails B | level_summaries (per-level duration_ms, correct_taps, wrong_taps), end_reason, total_taps, total_items | Per-level completion time (the standard TMT-B metric) and error classification. end_reason enables filtering out incomplete administrations. All legacy fields (correct_answers, score, percentageCorrectOverall, etc.) are retained. | Trails B docs |
Data collected before this update is not affected. These changes apply only to new activity completions. Your existing data remains valid and usable.
Rewritten Activities
The following activities have been substantially rewritten with new data schemas. If you have existing analysis pipelines for these games, review the updated documentation.
| Game | What Changed | Documentation |
|---|---|---|
| Balloon Risk (BART) | Rewritten with standard BART metrics. avg_pumps_collected is the primary risk-taking index (pumps on non-popped balloons only). balloon_results provides per-balloon pump counts and outcomes for learning-curve analysis. breakpoints echoes the hidden pop thresholds, enabling risk calibration analysis. The previous static_data.points field (total pumps from collected balloons) is now total_earnings (same value, same calculation). | Balloon Risk docs |
| Scratch Card | Rewritten with per-card results and structured data. avg_scratch_time_ms measures psychomotor speed. Per-card scratch_path (sampled pointer trajectory with timestamps) and scene element bounding boxes enable spatial analysis of visual attention and motor control. The previous version had minimal data output. | Scratch Card docs |
Post-Game Questionnaire
All modernized games now include a brief post-game questionnaire with two ratings (instruction clarity and willingness to repeat, each on a 1-5 scale). This data appears in the questionnaire field within static_data. Participants will see two emoji-based rating screens after completing each game.
Bug Fixes
| Game | Fix |
|---|---|
| Jewels A and B | Selecting "No" on the continue prompt was incorrectly advancing to the next level instead of ending the game. Now correctly ends the session. |
| Pop the Bubbles | Score screen now displays after the final round before the questionnaire. Improved tap responsiveness on mobile devices. |
| D-Cog | Dog visibility increased from 1 to 2 seconds. No-response rounds are no longer counted as successes. Audio feedback replaced with subtle chimes. |
| Symbol-Digit Substitution | Fixed an issue where the previous guess button stayed highlighted in the next round. |
| All games | Post-game questionnaire emoji faces now display in consistent left-to-right order (sad to happy). |
What This Means for You
Researchers with active studies: If you use Balloon Risk or Scratch Card, review the updated documentation as those data schemas have changed. For all other games, your existing analysis code will continue to work, since all legacy fields are preserved. New fields are available if you want richer data going forward.
Digital navigators and coordinators: The participant experience for existing games is largely unchanged. The most noticeable difference is the post-game questionnaire (two quick ratings after each game). No action is needed on your part.
Participants: Existing games look and feel the same. No action needed.
Reporting Issues
If you encounter any problems with the updated games, please email mindlamp@bidmc.harvard.edu.
Documentation
All activity documentation has been updated:
- Activity Reference: complete list of all activities
- Capabilities: Collect: updated showcase with game icons and cognitive domains