Open Cognitive in Strata on the web and choose Start Check-In. Complete the reaction and memory tasks, then answer the sharpness, mood and stress questions. Read each result with its own unit and label, and check the dated record afterward. These app tasks and self-ratings are not a clinical diagnosis or a substitute for assessment.
What happens in the reaction and memory tasks?
Select Begin on the Daily Check-In introduction. Reaction Time asks you to wait until the screen turns green, then tap. There are five recorded trials. An early tap produces a Too early message and retries the trial rather than recording that tap as a completed response.
Working Memory follows with a 2-Back task. Select Start after reading the instructions. Tap Match only when the current number matches the number shown exactly two positions earlier. In the constructed sequence 4, 9, 4, the final 4 matches the first one; matching the immediately preceding number is a different rule.
The current task presents 20 numbers. At level 2, scoring begins after the first two positions, leaving 18 scored decisions. Correctly leaving Match untouched on a nonmatching number counts as correct, so the accuracy result is not simply the percentage of times you pressed the button.
How should I answer the sharpness, mood and stress questions?
Mental Sharpness asks how mentally sharp you feel right now, using a choice from 1 to 10. This is your report of how you feel, separate from the timed and matching tasks. Keep that distinction when describing a session to someone else.
Read the endpoints displayed beneath the next questions. The web Mood scale labels 1 as Very Low and 10 as Amazing. The Stress Level scale labels 1 as Overwhelmed and 10 as At Peace. A higher stress-field number therefore corresponds to the calmer end of this particular scale, not more stress.
The final notes screen lets you add optional context. Describe an interruption, a misunderstanding of the task or another circumstance you want to remember. Save Check-In submits the session; Skip on this screen also submits, without requiring a note. It does not mean cancel the entire check-in.
What do milliseconds, accuracy and missing values mean?
The reaction result is a median in milliseconds after the app's trial filtering. It is not the fastest single tap. If too many trials are rejected, the app can leave the median absent and show Try again during the task. An absent result should not be rewritten as zero milliseconds.
The memory result is accuracy for the displayed task level. Keep that level alongside the percentage. A reaction time, an accuracy percentage and a subjective rating measure different things; adding them together does not create a meaningful combined result supplied by this workflow.
MedlinePlus explains that even clinical cognitive screening results do not establish a diagnosis on their own, and that concerns or online test results should be discussed with a health care provider. Its guidance does not validate Strata's tasks or provide normal ranges for them. Use the app record to describe the session rather than label yourself from a score.
Reference: MedlinePlus: Cognitive Testing
How do I check the saved date and avoid replacing a session?
Return to Cognitive and inspect the selected day's check-in and Check-In History. Match the date and displayed values to the session you intended to record. The completion screen is a useful prompt to do this check, rather than independent proof that every field has persisted.
The web check-in saves against the current local date. It does not offer a past-date selector inside the task flow. Its daily save updates the existing entry for the same user and date, so repeating and saving on that date can replace earlier values rather than create a separate trial session in history.
Before comparing days, keep the task, level and any recorded context visible. If you were interrupted, say that in the record instead of deciding from the number alone that your cognition changed. The guide explains how to preserve and read observations, not how to diagnose a change or improve a clinical outcome.
- Review the completed session's labels and units.
- Open Cognitive and select the intended date.
- Check the dated entry and history before relying on the result.
- Consider the replacement behavior before saving another check-in on the same day.
An illustrative example
What does the memory percentage describe?
In a fictional level-2 session, Alex makes 16 correct decisions across the 18 scored positions. The app's rounding produces 89% accuracy. That describes this constructed matching task, not an intelligence score, a clinical screening result or a diagnosis. A note that the instructions were reread preserves context without supplying a medical explanation.
What other questions come up?
Does a faster tap prove better overall cognition?
The recorded reaction result describes performance on this app task. It does not establish an overall clinical assessment. Keep the task definition and session context with the value.
Can I use a zero for an invalid reaction result?
No. A missing median means the app did not retain a usable median for that session. Zero is a numeric value with a different meaning and would misrepresent the record.
Are the mood and stress numbers interchangeable?
They answer different questions and have different endpoint labels. Preserve the field name when copying a rating rather than using an unlabeled number out of ten.
Is a cognitive activity log the same as this check-in?
No. The Cognitive page also has activity records, such as reading or learning entries. Those logs describe activities; the check-in records the task results and self-ratings from this sequence.
Keep this in mind
Keep the task result, self-rating, date and session context distinct. Verify the record before comparing sessions, and take clinical questions to an appropriate professional.
Strata Health Vault
Put this guide into practice with your own health data.
Bring sleep, movement, meals, and vitals into one unified vault, and connect your data to Claude, ChatGPT, or custom AI assistants.