NeuroFlex Behavioral Event Model v1
===================================

Document purpose
----------------

This document defines the first version of the NeuroFlex behavioral event model.

The goal is not to turn the MVP into a research platform immediately.
The goal is to design the MVP so that every activity can later produce structured behavioral data without redesigning the application from scratch.

NeuroFlex Phase 1 may remain:
- free
- offline
- without login
- without backend
- without AI
- without cloud database

However, the app should be designed as if a backend may exist later.

Current MVP:
saveLocally(event)

Future research platform:
saveLocally(event)
syncToCloud(event)

Core product direction
----------------------

NeuroFlex is not only a brain training app.

The long-term direction is to become a longitudinal behavioral observatory for cognitive health and human behavior.

The app should observe behavior across:
- cognition
- memory
- attention
- reaction speed
- emotional state
- motivation
- loneliness
- social connection
- wellness habits
- motor function
- routine adherence
- avoidance behavior
- persistence
- self-perception
- behavioral change over time

The long-term research objective is not:
"Diagnose Alzheimer's disease from app usage."

The stronger and safer objective is:
"Characterize long-term behavioral trajectories and investigate whether specific trajectories are associated with future cognitive outcomes."

Core data philosophy
--------------------

NeuroFlex should never store only final results.

Weak data:
score = 84

Useful data:
- score
- reaction times
- time to first action
- hesitation time
- corrections
- mistakes
- retries
- abandonment
- difficulty level
- session duration
- interaction sequence
- change over time

The process may be more valuable than the outcome.

Two users can achieve the same score while showing very different behavioral patterns.

Example:
User A:
- fast responses
- stable reaction times
- few corrections

User B:
- same score
- slower responses
- more hesitation
- more corrections
- higher reaction time variability

For future analysis, these are different behavioral profiles.

General event envelope
----------------------

Every event should follow the same base structure.

Example:

{
  "event_id": "uuid",
  "anonymous_user_id": "uuid",
  "session_id": "uuid",
  "timestamp_utc": "2026-06-01T10:00:00Z",
  "app_version": "2.4",
  "event_type": "memory_card_selected",
  "event_category": "cognitive",
  "screen": "memory_cards",
  "activity_id": "memory_cards",
  "sequence_index": 14,
  "payload": {}
}

Required fields
---------------

event_id
- Unique identifier for the event.

anonymous_user_id
- Random app-generated user identifier.
- Not email.
- Not name.
- Not phone number.
- Not device advertising ID.

session_id
- Unique ID for one app session.
- New session starts after app open or after a long inactivity window.

timestamp_utc
- Event time in UTC.

app_version
- Version of NeuroFlex.

event_type
- Specific event name.

event_category
- High-level category such as system, cognitive, emotional, wellness, social, motor, behavioral.

screen
- Current app screen.

activity_id
- Activity or feature where event occurred.

sequence_index
- Incrementing number within the session to reconstruct event order.

payload
- Event-specific data.

Recommended common payload fields
---------------------------------

{
  "local_time": "08:42",
  "timezone_offset_minutes": 120,
  "device_orientation": "portrait",
  "input_method": "touch",
  "difficulty_level": 2,
  "is_practice": false,
  "was_completed": true,
  "duration_ms": 42000
}

Privacy principles
------------------

Do not collect by default:
- name
- email
- phone number
- precise location
- contacts
- microphone recordings
- camera images
- medical records
- free-text health information
- government IDs
- identity documents

Optional future research data must require explicit informed consent.

Diagnosis-related data, if ever added, should be:
- optional
- user-controlled
- clearly explained
- separated from normal wellness usage
- handled as sensitive research data

Data tiers
----------

NeuroFlex should separate event collection into tiers.

Tier 1 - Core MVP events
Always safe and useful.
Can be stored locally even in offline MVP.

Examples:
- activity started
- activity completed
- activity abandoned
- score
- difficulty level
- reaction time
- error count
- duration
- task completion
- check-in values

Tier 2 - Detailed behavioral research events
Useful for future research mode.
Should be optional or clearly explained.

Examples:
- full interaction sequence
- hesitation time per decision
- time between taps
- correction frequency
- navigation path
- repeated retries
- detailed notification response behavior

Tier 3 - Advanced motor and high-frequency data
Potentially valuable but heavier and more sensitive.
Should only be enabled in explicit research mode.

Examples:
- full touch paths
- drawing trajectories
- tremor-like finger movement
- line tracing paths
- spiral drawing raw data
- high-frequency tap timing

Event categories
================

Level 0 - System Events
-----------------------

app_install
Purpose:
First known app installation/opening event.

Payload:
{
  "install_source": "unknown",
  "first_app_version": "2.4"
}

app_open
Purpose:
User opens the app.

Payload:
{
  "open_source": "launcher|notification|deep_link|unknown"
}

app_close
Purpose:
App is closed or moved to background.

Payload:
{
  "session_duration_ms": 120000
}

session_start
Purpose:
Start of a behavioral session.

Payload:
{
  "start_reason": "app_open|return_from_background|notification"
}

session_end
Purpose:
End of a behavioral session.

Payload:
{
  "duration_ms": 120000,
  "screens_viewed_count": 4,
  "activities_started_count": 2,
  "activities_completed_count": 1
}

app_inactive_timeout
Purpose:
User became inactive for a defined period.

Payload:
{
  "inactive_duration_ms": 300000,
  "last_screen": "home"
}

app_version_changed
Purpose:
App version changed after update.

Payload:
{
  "previous_version": "2.3",
  "new_version": "2.4"
}

Level 1 - Navigation and Interaction Events
-------------------------------------------

screen_view
Purpose:
User views a screen.

Payload:
{
  "screen": "home",
  "previous_screen": "welcome",
  "entry_source": "button|notification|back_navigation|auto"
}

screen_exit
Purpose:
User leaves a screen.

Payload:
{
  "screen": "home",
  "duration_ms": 35000,
  "exit_target": "games",
  "exit_type": "button|back|app_close|timeout"
}

button_press
Purpose:
User presses a meaningful UI button.

Payload:
{
  "button_id": "btnGameMemoryCards",
  "button_label": "Memory Cards",
  "time_since_screen_view_ms": 2300
}

back_navigation
Purpose:
User goes back.

Payload:
{
  "from_screen": "settings",
  "to_screen": "home"
}

menu_opened
Purpose:
User opens a menu.

Payload:
{
  "menu_id": "top_menu",
  "screen": "home"
}

menu_item_selected
Purpose:
User selects a menu item.

Payload:
{
  "menu_id": "top_menu",
  "item_id": "permissions"
}

repeated_tap
Purpose:
Detects repeated taps on same area or control.

Payload:
{
  "target_id": "submit_button",
  "tap_count": 3,
  "window_ms": 1200
}

invalid_tap
Purpose:
Tap outside active target during a task.

Payload:
{
  "x": 214,
  "y": 812,
  "nearest_target_id": "card_4",
  "distance_to_target_center_px": 82
}

Level 2 - Activity Lifecycle Events
-----------------------------------

activity_start
Purpose:
Any game, wellness activity, check-in, or task starts.

Payload:
{
  "activity_id": "memory_cards",
  "activity_category": "train",
  "difficulty_level": 2
}

activity_complete
Purpose:
Activity is completed.

Payload:
{
  "activity_id": "memory_cards",
  "duration_ms": 65000,
  "difficulty_level": 2,
  "completion_rate": 1.0
}

activity_abandoned
Purpose:
Activity started but not completed.

Payload:
{
  "activity_id": "memory_cards",
  "duration_ms": 18000,
  "progress_percent": 35,
  "abandon_reason": "back|home|app_close|timeout|unknown"
}

activity_skipped
Purpose:
User intentionally skips activity.

Payload:
{
  "activity_id": "daily_checkin",
  "skip_source": "button",
  "time_since_prompt_ms": 4300
}

activity_retry
Purpose:
User retries activity.

Payload:
{
  "activity_id": "sequence_memory",
  "retry_count_today": 2,
  "previous_result": "failed|completed|abandoned"
}

activity_returned
Purpose:
User returns to a previously abandoned activity.

Payload:
{
  "activity_id": "number_sequence",
  "time_since_abandon_ms": 86400000
}

Level 3 - Cognitive Events
--------------------------

Memory Cards
------------

memory_cards_round_start
Payload:
{
  "level": 2,
  "grid_rows": 4,
  "grid_columns": 4,
  "pair_count": 8
}

memory_card_selected
Payload:
{
  "card_position": 5,
  "card_symbol_id": "apple",
  "selection_index": 1,
  "time_since_round_start_ms": 2200,
  "time_since_previous_selection_ms": 1300
}

memory_pair_attempt
Payload:
{
  "first_card_position": 5,
  "second_card_position": 9,
  "is_match": false,
  "time_between_cards_ms": 1800,
  "hesitation_ms": 1800
}

memory_pair_matched
Payload:
{
  "pair_symbol_id": "apple",
  "attempt_count_for_pair": 2,
  "time_since_round_start_ms": 18000
}

memory_cards_round_end
Payload:
{
  "level": 2,
  "completed": true,
  "duration_ms": 62000,
  "attempt_count": 14,
  "wrong_pair_count": 6,
  "correct_pair_count": 8,
  "reshuffle_count": 0,
  "average_pair_decision_ms": 2100,
  "reaction_time_variability_ms": 620
}

Sequence Memory
---------------

sequence_round_start
Payload:
{
  "level": 3,
  "sequence_length": 5,
  "button_count": 4
}

sequence_display_start
Payload:
{
  "sequence_length": 5,
  "display_speed_ms": 900
}

sequence_user_input
Payload:
{
  "input_index": 2,
  "selected_button_id": "blue",
  "expected_button_id": "red",
  "is_correct": false,
  "time_since_input_start_ms": 3400,
  "time_since_previous_input_ms": 900
}

sequence_round_end
Payload:
{
  "level": 3,
  "completed": false,
  "first_error_position": 2,
  "correct_prefix_length": 2,
  "duration_ms": 12000,
  "retry_count": 1
}

Missing Item
------------

missing_item_memorize_start
Payload:
{
  "level": 2,
  "item_count": 6,
  "memorize_duration_ms": 7000
}

missing_item_choice
Payload:
{
  "missing_item_id": "banana",
  "selected_item_id": "orange",
  "is_correct": false,
  "time_to_choice_ms": 3200,
  "option_count": 3
}

missing_item_round_end
Payload:
{
  "level": 2,
  "completed": true,
  "attempt_count": 2,
  "duration_ms": 15000
}

Number Sequence
---------------

number_sequence_display
Payload:
{
  "level": 2,
  "sequence_length": 5,
  "display_duration_ms": 5000
}

number_sequence_typing_start
Payload:
{
  "time_since_hidden_ms": 1200
}

number_sequence_input_changed
Payload:
{
  "input_length": 3,
  "backspace_count": 1,
  "time_since_typing_start_ms": 2500
}

number_sequence_submit
Payload:
{
  "sequence_length": 5,
  "is_correct": false,
  "error_type": "substitution|omission|insertion|order_swap|unknown",
  "first_error_position": 3,
  "typing_duration_ms": 4100,
  "backspace_count": 2,
  "submit_latency_ms": 900
}

Reaction Test
-------------

reaction_test_start
Payload:
{
  "trial_count": 10,
  "stimulus_type": "visual"
}

reaction_trial
Payload:
{
  "trial_index": 4,
  "reaction_ms": 341,
  "is_valid": true,
  "false_start": false
}

reaction_test_end
Payload:
{
  "trial_count": 10,
  "valid_trial_count": 9,
  "average_reaction_ms": 380,
  "median_reaction_ms": 361,
  "reaction_variability_ms": 74,
  "false_start_count": 1
}

Pattern Matching
----------------

pattern_task_start
Payload:
{
  "level": 2,
  "pattern_type": "shape|color|number|mixed",
  "option_count": 4
}

pattern_attempt
Payload:
{
  "is_correct": true,
  "time_to_answer_ms": 2100,
  "selected_option_id": "option_2",
  "correct_option_id": "option_2"
}

Visual Search / Attention
-------------------------

attention_task_start
Payload:
{
  "target_type": "star",
  "distractor_count": 18,
  "target_count": 1
}

attention_target_tap
Payload:
{
  "is_correct_target": true,
  "time_to_target_ms": 2900,
  "distance_from_target_center_px": 12
}

attention_task_end
Payload:
{
  "duration_ms": 14000,
  "correct_targets": 5,
  "missed_targets": 1,
  "wrong_taps": 2
}

Recall Tasks
------------

recall_prompt_shown
Payload:
{
  "prompt_type": "word|image|number|story",
  "item_count": 4,
  "delay_ms": 30000
}

recall_attempt
Payload:
{
  "response_type": "typed|selected|spoken_not_recorded",
  "is_correct": true,
  "time_to_first_action_ms": 1800,
  "response_duration_ms": 4300,
  "correction_count": 1
}

Level 4 - Emotional and Self-Assessment Events
----------------------------------------------

daily_checkin_start
Payload:
{
  "checkin_type": "daily",
  "prompt_version": "v1"
}

mood_check
Payload:
{
  "value": "excellent|good|calm|motivated|lonely|stressed|tired|sad",
  "scale_value": 4,
  "time_to_answer_ms": 2600
}

energy_check
Payload:
{
  "value": 3,
  "scale_min": 1,
  "scale_max": 5,
  "time_to_answer_ms": 1700
}

motivation_check
Payload:
{
  "value": 2,
  "scale_min": 1,
  "scale_max": 5
}

focus_check
Payload:
{
  "value": 4,
  "scale_min": 1,
  "scale_max": 5
}

stress_check
Payload:
{
  "value": 2,
  "scale_min": 1,
  "scale_max": 5
}

loneliness_check
Payload:
{
  "value": 3,
  "scale_min": 1,
  "scale_max": 5
}

confidence_check
Payload:
{
  "activity_id": "memory_cards",
  "value": 4,
  "scale_min": 1,
  "scale_max": 5
}

perceived_difficulty
Payload:
{
  "activity_id": "sequence_memory",
  "value": 3,
  "scale_min": 1,
  "scale_max": 5
}

memory_self_rating
Payload:
{
  "value": 3,
  "scale_min": 1,
  "scale_max": 5
}

sleep_quality_check
Payload:
{
  "value": 4,
  "scale_min": 1,
  "scale_max": 5
}

Level 5 - Wellness Pillar Events
--------------------------------

Train
-----

train_opened
Payload:
{
  "entry_source": "home|bottom_nav|reminder"
}

train_activity_selected
Payload:
{
  "activity_id": "memory_cards",
  "selection_time_ms": 1900
}

Drink
-----

drink_reminder_scheduled
Payload:
{
  "scheduled_time": "10:00",
  "reminder_type": "hydration"
}

drink_reminder_received
Payload:
{
  "notification_id": "uuid"
}

drink_completed
Payload:
{
  "amount_ml": 250,
  "completion_source": "notification|manual|schedule"
}

drink_skipped
Payload:
{
  "reason": "not_now|dismissed|unknown",
  "time_since_reminder_ms": 5000
}

Move
----

move_activity_start
Payload:
{
  "activity_id": "seated_stretch",
  "activity_type": "stretching|walking|balance|mobility|custom",
  "planned_duration_minutes": 5
}

move_activity_complete
Payload:
{
  "activity_id": "seated_stretch",
  "actual_duration_minutes": 5,
  "perceived_difficulty": 2
}

Eat
---

eat_activity_logged
Payload:
{
  "activity_type": "healthy_meal|snack|meal_planning|custom",
  "time_to_log_ms": 2400
}

healthy_habit_complete
Payload:
{
  "habit_id": "eat_breakfast",
  "completion_source": "manual|reminder|schedule"
}

Connect
-------

connect_prompt_shown
Payload:
{
  "prompt_type": "call_someone|message_someone|visit_someone|gratitude|custom"
}

connect_completed
Payload:
{
  "prompt_type": "call_someone",
  "completion_source": "manual",
  "meaningfulness_rating": 4
}

connect_skipped
Payload:
{
  "prompt_type": "call_someone",
  "time_to_skip_ms": 3100
}

social_contact_logged
Payload:
{
  "contact_type": "call|message|in_person|group|other",
  "duration_category": "short|medium|long|unknown",
  "meaningfulness_rating": 4
}

Relax
-----

relax_activity_start
Payload:
{
  "activity_id": "breathing_4_4",
  "activity_type": "breathing|meditation|body_scan|sleep|custom"
}

relax_activity_complete
Payload:
{
  "activity_id": "breathing_4_4",
  "duration_ms": 180000,
  "calm_after_rating": 4
}

relax_activity_abandoned
Payload:
{
  "activity_id": "breathing_4_4",
  "duration_ms": 42000,
  "progress_percent": 25
}

Level 6 - Schedule, Tasks, and Routine Events
---------------------------------------------

task_created
Payload:
{
  "task_category": "train|drink|move|eat|connect|relax|custom",
  "has_reminder": true,
  "repeat_type": "none|daily|weekly|custom"
}

task_edited
Payload:
{
  "task_id": "uuid",
  "changed_fields": ["time", "repeat_type"]
}

task_deleted
Payload:
{
  "task_id": "uuid",
  "task_age_days": 12
}

task_due
Payload:
{
  "task_id": "uuid",
  "scheduled_time": "09:00"
}

task_completed
Payload:
{
  "task_id": "uuid",
  "task_category": "drink",
  "time_since_due_ms": 600000,
  "completion_source": "notification|schedule|manual"
}

task_postponed
Payload:
{
  "task_id": "uuid",
  "postpone_minutes": 30,
  "postpone_count_today": 2
}

task_skipped
Payload:
{
  "task_id": "uuid",
  "time_since_due_ms": 1200000,
  "skip_source": "manual|timeout|unknown"
}

task_ignored
Payload:
{
  "task_id": "uuid",
  "ignored_duration_ms": 86400000
}

routine_day_summary
Payload:
{
  "date": "2026-06-01",
  "planned_task_count": 8,
  "completed_task_count": 5,
  "skipped_task_count": 1,
  "postponed_task_count": 3,
  "completion_rate": 0.625
}

Level 7 - Notification Events
-----------------------------

notification_scheduled
Payload:
{
  "notification_id": "uuid",
  "notification_type": "drink|move|task|checkin|game|relax",
  "scheduled_time": "10:00"
}

notification_received
Payload:
{
  "notification_id": "uuid",
  "notification_type": "drink",
  "delivered": true
}

notification_opened
Payload:
{
  "notification_id": "uuid",
  "time_since_received_ms": 180000
}

notification_dismissed
Payload:
{
  "notification_id": "uuid",
  "time_since_received_ms": 12000
}

notification_ignored
Payload:
{
  "notification_id": "uuid",
  "ignored_duration_ms": 3600000
}

notification_action_selected
Payload:
{
  "notification_id": "uuid",
  "action_id": "complete|snooze|skip|open"
}

Level 8 - Behavioral Pattern Events
-----------------------------------

challenge_avoided
Purpose:
User avoids or exits a challenge before meaningful engagement.

Payload:
{
  "activity_id": "reaction_test",
  "time_before_exit_ms": 2800,
  "previous_failures_count": 2
}

difficulty_decreased
Payload:
{
  "activity_id": "memory_cards",
  "previous_level": 4,
  "new_level": 3,
  "reason": "user_selected|auto_adjusted"
}

difficulty_increased
Payload:
{
  "activity_id": "memory_cards",
  "previous_level": 2,
  "new_level": 3,
  "reason": "user_selected|auto_adjusted"
}

long_pause_detected
Payload:
{
  "activity_id": "number_sequence",
  "pause_duration_ms": 8000,
  "pause_context": "before_answer|during_typing|after_error"
}

return_after_gap
Payload:
{
  "days_since_last_session": 5,
  "return_source": "manual|notification"
}

preference_signal
Payload:
{
  "preferred_activity_id": "memory_cards",
  "avoided_activity_id": "reaction_test",
  "basis": "selection_frequency|completion_rate|manual_favorite"
}

Level 9 - Motor and Touch Events
--------------------------------

These should generally be Tier 2 or Tier 3, not always-on MVP data.

tap_event
Payload:
{
  "x": 120,
  "y": 620,
  "target_id": "button_submit",
  "target_center_x": 132,
  "target_center_y": 612,
  "distance_from_target_center_px": 14,
  "time_to_target_ms": 2100
}

touch_down
Payload:
{
  "x": 120,
  "y": 620,
  "pressure": null,
  "target_id": "card_4"
}

touch_up
Payload:
{
  "x": 122,
  "y": 618,
  "duration_ms": 120,
  "target_id": "card_4"
}

drag_start
Payload:
{
  "x": 180,
  "y": 700,
  "target_id": "slider"
}

drag_end
Payload:
{
  "duration_ms": 1400,
  "path_point_count": 32,
  "distance_px": 240,
  "correction_count": 3
}

touch_path
Payload:
{
  "activity_id": "line_trace",
  "points": [
    {"t": 0, "x": 100, "y": 500},
    {"t": 16, "x": 104, "y": 502}
  ],
  "duration_ms": 3200
}

hold_still_test
Payload:
{
  "duration_ms": 10000,
  "movement_radius_avg_px": 4.2,
  "movement_radius_max_px": 11.7,
  "stability_score": 0.82
}

line_trace_task
Payload:
{
  "duration_ms": 12000,
  "average_deviation_px": 8.4,
  "max_deviation_px": 32.1,
  "correction_count": 12,
  "smoothness_score": 0.76
}

spiral_drawing_task
Payload:
{
  "duration_ms": 18000,
  "raw_path_available": true,
  "average_deviation_px": 9.2,
  "smoothness_score": 0.71,
  "tremor_proxy_score": 0.18
}

Level 10 - Research Consent and Data Governance Events
------------------------------------------------------

research_consent_shown
Payload:
{
  "consent_version": "research_v1",
  "language": "en"
}

research_consent_accepted
Payload:
{
  "consent_version": "research_v1",
  "accepted_at_utc": "2026-06-01T10:00:00Z"
}

research_consent_declined
Payload:
{
  "consent_version": "research_v1"
}

research_consent_revoked
Payload:
{
  "consent_version": "research_v1",
  "revoked_at_utc": "2026-06-01T10:00:00Z"
}

data_export_requested
Payload:
{
  "request_source": "settings"
}

local_data_deleted
Payload:
{
  "delete_source": "settings",
  "event_count_deleted": 1240
}

Derived metrics
===============

The app should collect raw events first.
Derived metrics can be calculated later.

Potential derived metrics:

Session metrics
- sessions per day
- average session duration
- session fragmentation
- days active per week
- longest inactivity gap

Cognitive metrics
- average reaction time
- reaction time variability
- learning curve slope
- retention curve
- first-error position
- correction rate
- completion time trend

Behavioral metrics
- task adherence rate
- postponement frequency
- skip frequency
- abandonment rate
- return-after-abandonment rate
- challenge avoidance pattern

Emotional metrics
- mood trend
- energy trend
- loneliness trend
- self-rated memory trend
- difference between confidence and objective performance

Motor metrics
- tap precision
- touch stability
- tracing deviation
- movement smoothness
- correction frequency

Circadian metrics
- preferred usage time
- usage time drift
- routine stability
- night usage frequency
- morning engagement frequency

Recommended MVP implementation
==============================

For the current offline MVP, implement a local event logger with no cloud sync.

Minimum local architecture:

EventLogger.log(eventType, category, screen, activityId, payload)

Storage:
- local JSON lines file, or
- local SQLite table, or
- local Room database

Recommended local fields:
- event_id
- anonymous_user_id
- session_id
- timestamp_utc
- app_version
- event_type
- event_category
- screen
- activity_id
- sequence_index
- payload_json

The MVP does not need to expose this data to users immediately.
However, for testing and development, it may be useful to add:
- export local event log
- clear local event log
- show event count

MVP priority events
-------------------

If implementation must be limited, start with these:

System:
- app_open
- session_start
- session_end
- screen_view
- screen_exit

Activity:
- activity_start
- activity_complete
- activity_abandoned
- activity_skipped

Games:
- round_start
- user_action
- mistake
- correction
- round_end

Tasks:
- task_created
- task_completed
- task_postponed
- task_skipped
- task_ignored

Check-ins:
- mood_check
- energy_check
- focus_check
- stress_check
- loneliness_check
- memory_self_rating

Notifications:
- notification_received
- notification_opened
- notification_dismissed
- notification_action_selected

Design rules for future activities
==================================

Every new activity should define:

1. Start event
2. Step/action events
3. Error events
4. Correction events
5. Completion event
6. Abandonment event
7. Difficulty metadata
8. Timing metadata
9. Optional confidence/difficulty self-rating

Every game should measure:
- time to first action
- time between actions
- response time
- error count
- correction count
- retry count
- completion time
- abandonment point
- difficulty level

Every wellness activity should measure:
- prompt shown
- activity opened
- activity completed
- activity skipped
- time between prompt and action
- perceived difficulty where relevant

Every check-in should measure:
- value selected
- time to answer
- skipped or completed
- change compared with previous check-ins

Scientific language guidelines
==============================

Use careful wording.

Avoid:
- "NeuroFlex diagnoses Alzheimer's."
- "NeuroFlex predicts dementia."
- "AI detects disease."
- "This app prevents Alzheimer's."

Prefer:
- "NeuroFlex collects behavioral data for future research."
- "NeuroFlex characterizes behavioral trajectories."
- "NeuroFlex investigates associations between behavioral patterns and future cognitive outcomes."
- "NeuroFlex may support research into digital behavioral biomarkers."

Important methodological note
=============================

Do not frame the future dataset as simply:
diagnosed vs not diagnosed.

This is scientifically weak because "not diagnosed" does not mean "healthy."

Some users may:
- have preclinical disease
- have mild cognitive impairment
- receive a diagnosis years later
- have other neurological or psychological factors

The better research approach:

1. Collect broad longitudinal behavioral data.
2. Identify natural behavioral trajectories and clusters.
3. Study how these trajectories evolve over time.
4. Later investigate whether specific trajectories are associated with clinical outcomes.

The initial machine learning objective should be discovery, not prediction.

Potential future ML methods:
- clustering
- anomaly detection
- trajectory modeling
- latent variable discovery
- temporal pattern analysis
- population segmentation
- resilience modeling

Final principle
===============

NeuroFlex should collect behavioral signals not because we already know which signals matter, but because we do not yet know which signals may become meaningful after years of observation.

The long-term value of NeuroFlex depends on:
- breadth of behavioral observation
- depth of interaction data
- consistency of event structure
- duration of longitudinal data
- scientific caution
- privacy-first design

Brain games are only one part of NeuroFlex.

The real product is the structured observation of human behavior over time.