Compare Two Wearable Exports With a Separate Personal Log

Freeze two dated exports and confirm ownership

Save each export or screenshot with its original filename, capture time, device model, app, account, date label, and time zone. Confirm that both records belong to the same intended wearer and that neither was created during a shared-device or reset interval. Do not edit values, crop away missing-data messages, or replace the earlier export with a newly recalculated view. If only a summary tile exists for one date and a detailed export for the other, record that source difference before proceeding. The first requirement is two stable artifacts, not two perfectly matched charts.

Use one field table for both device records

Create identical columns for displayed interval, stage labels, durations, percentages, score if present, untracked time, missing symbols, software version, and export source. Copy product vocabulary exactly on each date. Do not force a category that appears on only one export into the other; mark the absent field as not displayed. Keep units and rounding as shown. When totals disagree with the chart interval, record each published value and the visible difference. A shared table makes mismatches readable without pretending the device measured a single fixed set of fields.

Attach one independent personal log to each date

Prepare a separate factual log for each overnight interval using the same personal fields: approximate bed entry, getting up, deliberate device removal, clock checks, planned alarm, and known interruption. Label the source of every time and use ranges for later memory. Do not use either export to create events that were not independently known. Likewise, do not erase a remembered event because the device chart lacks a matching label. The two personal logs support timestamp comparison only; they do not certify the stage estimates or turn recall into sensor data.

List setup and capture differences before comparing values

Check software and app versions, tracking settings, device placement as ordinarily recorded, charging or removal, account time zone, export time, and visible gaps. Place every difference in a setup block above the value table. If a change makes the fields incompatible, keep the comparison but narrow it to fields that truly match, or return unknown. Do not normalize one export by guessing how another version would have labeled it. Missing setup facts are not small details; they define the boundary of what the two device records can show.

Describe value and timestamp differences without ranking

Compare displayed categories one row at a time, then compare the device intervals with their respective personal-log intervals. State clock differences and missing regions in neutral terms. A higher product score, longer displayed stage, or closer timestamp match is simply a larger displayed number or smaller offset on that product’s scale. Do not call either date better, accurate, successful, or failed. Do not average the two dates into a personal baseline. Preserve opposing details, such as a longer device interval paired with more missing segments, instead of choosing the one that supports a preferred story.

Return observed, mixed, or unknown and archive the pair

Use observed when matching fields and setup facts permit a precise statement about similarities or differences. Use mixed when the exports move in several directions or the device and personal-log offsets differ by field. Use unknown when ownership, settings, dates, capture gaps, or comparable labels are unresolved. Write one closing sentence naming the device, both dates, the principal displayed difference, and the largest evidence boundary. Add no threshold, brand ranking, future prediction, or automatic instruction. Archive the two exports, their logs, and the table, then stop.

Keep exploring this topic