Activity Data Format

All activities produce ActivityEvents — timestamped records containing session summaries and per-action detail.

ActivityEvent Structure

{
  "timestamp": 1634047106612,
  "activity": "e2wxcekbb0typzdtz0pb",
  "duration": 3735,
  "static_data": { ... },
  "temporal_slices": [
    { "item": ..., "value": ..., "type": ..., "duration": ..., "level": ... }
  ]
}

Field	Type	Description
timestamp	integer	Unix timestamp in milliseconds when the activity was completed
activity	string	The activity's unique identifier
duration	integer	Total time spent on the activity in milliseconds
static_data	object	Session-level summary (overall scores, points, metadata)
temporal_slices	array	Per-action detail (each question answered, each tap made, each level completed)

temporal_slices

Each element represents a single action — a survey question answered, a game level completed, or a tap recorded.

Field	Type	Description
item	string or number	Question text (surveys), tap/pump number (games), or level identifier
value	any	The response — selected option (surveys), score (games), correctness flag
type	any	Question type (surveys), correctness boolean (games), or null
duration	number	Time spent on this item in milliseconds
level	number or null	Game level number, or null for surveys

The meaning of each field varies by activity type. See the individual activity pages in Activity Reference for per-activity schemas.

Data Retrieval

Data can be retrieved through:

• REST API — Query activity events for specific participants, time ranges, and types. See the API documentation.
• Python SDK — import LAMP for programmatic data access. See Python SDK.
• Cortex — Process raw events into computed behavioral features automatically. See Cortex.
• Dashboard — View and export data through the Data Portal or the Portal tab.

Cortex Analysis

The Cortex pipeline transforms raw activity events into behavioral insights through three stages:

1. Raw features — Retrieve activity event data directly (e.g., survey, jewels_a, balloon_risk).
2. Primary features — Compute per-session metrics (e.g., survey_scores, game_level_scores).
3. Secondary features — Aggregate scores across time windows (e.g., survey_results, game_results).

See the Cortex documentation for details on running activity analysis. For per-activity schemas, see the Activity Reference.