What this page establishes: two CSV files are available below, and their row counts and numerical summaries were checked on September 17, 2026. Checking a spreadsheet’s arithmetic does not independently verify participant recruitment, measurements, consent, or the origin of its observations. Do not treat these files as clinical evidence.
Available files and verification status
The files have previously been described on this site as an internal journaling pilot and an affirmation-audio beta. That description is a site-reported provenance claim, not independent validation. The checks below are deliberately narrower: what the downloadable files contain and what can be calculated from them.
Journaling pilot CSV: 71 data rows
Download the journaling pilot CSV. The file contains participant labels, self-reported demographic categories, total nights journaled out of 14, two cortisol-strip columns, and two self-tracked task-count columns. The notes identify consumer-grade strips, self-report, and no control group.
Recalculated column means:
- Cortisol-strip column, baseline: 9.4014; day 10: 7.1944, in the units labeled nmol/L in the file.
- Task-count column, baseline: 3.0986; day 10: 3.7028.
- Total nights journaled: 7.8451 out of 14, averaged across the 71 rows.
These are arithmetic summaries of the supplied columns. The cortisol-strip columns should not be interpreted as validated laboratory measurements. A before/after difference without a comparison group does not establish that journaling caused the difference.
Affirmation-audio beta CSV: 27 data rows
Download the affirmation-audio beta CSV. The file contains participant labels, baseline and day-14 self-assessed annoyance ratings, playback adjustments, days used out of 14, and notes.
- Mean baseline annoyance rating: 6.3111.
- Mean day-14 annoyance rating: 3.1519.
- Mean days used: 9.4444 out of 14.
The columns label annoyance on a 1–10 scale; the file notes describe the assessment as unvalidated. These values summarize that file, not a validated psychological outcome or a recommended treatment.
Corrections and limits readers should know
- Affirmation means: the methodology page previously reported a 6.2 → 2.9 change. Those values do not match the current CSV’s column means of approximately 6.31 → 3.15. Use the explicitly calculated values above when describing this file; do not cite the older summary as a verified result.
- Journaling adherence: this page previously reported that 68% of participants journaled at least five nights in seven. The CSV records only a total across 14 nights, not the nightly sequence. That weekly adherence claim cannot be reproduced from the supplied columns and has been removed from this page.
- Provenance: a public file and plausible averages cannot establish whether the underlying observations were collected as described. No independent participant-level provenance verification is claimed here.
- Privacy: row labels are not a guarantee of anonymization. Do not attempt to identify people represented by a dataset or combine demographic fields to infer identity.
Reproduce the arithmetic in a spreadsheet
Download the file and open it as CSV. Keep the header row. Count the data rows, then use AVERAGE on the relevant numeric column:
- Journaling baseline cortisol:
=AVERAGE(E2:E72); day 10:=AVERAGE(F2:F72). - Journaling baseline tasks:
=AVERAGE(G2:G72); day 10:=AVERAGE(H2:H72). - Affirmation baseline rating:
=AVERAGE(B2:B28); day 14:=AVERAGE(C2:C28).
If the download changes, the row ranges and answers may also change. Compare the file’s headers before applying a formula. Check missing values, duplicate labels, and the measurement definitions before making further comparisons.
Research, internal observations, and illustrations are different
Published studies belong to their named authors. A site-reported internal observation is not equivalent to a peer-reviewed study. An illustrative example is not a measured result at all.
Our methodology register identifies some figures elsewhere on the site as illustrative composites, including “73% of 2,847 participants” and “1,300 users.” They must not be presented or cited as datasets verified by this page. A link to this register cannot substitute for labeling an illustration at the point where it appears.
For a concrete example of published evidence with an explicit design and limitations, see Albulescu and colleagues’ 2022 micro-break meta-analysis. It reports benefits for vigor and fatigue, but not a statistically significant overall performance effect. A research citation should support the specific sentence attached to it, not merely concern the same broad topic.
Reuse and corrections
The previous version of this page described the internal CSVs as CC BY 4.0. A reuse label does not establish participant consent, data accuracy, or suitability for a new purpose. Check the underlying provenance and any applicable privacy obligations before redistribution. Third-party datasets remain under their respective owners’ terms.
If a published article conflicts with a downloaded file, report the article URL, file name, column, and calculation. Do not send personal participant information. See the editorial policy for the site’s stated correction standards.