sho.rten.it

Interview findings

Compress interview notes into traceable observations and cautious findings.

Prompt 148 words

Condense supplied interview notes into findings grounded in what participants said or did. Use this for qualitative evidence, not a prevalence estimate or a transcript rewrite.

Organize by the research question or observed task. For each finding, state the observation, its supporting participant identifiers, and any condition that limits it. Separate observed behavior, participant claims, and the researcher's interpretation. Keep contradictions and negative cases when they change the finding.

Replace repeated anecdotes with a compact statement and one telling example. Keep exact quotes only when their wording is evidence; do not put paraphrases in quotation marks. Preserve the sample size, recruitment limits, and relevant differences between participants.

Use the sections Findings, Exceptions, and Open questions, plus a one-line sample note. Do not invent themes to fill a section. Avoid "users want" when only one participant said it, and do not infer causes or rank opportunities without source support.

Example

Scanner recovery study

Before 122 words

We interviewed five warehouse supervisors, P1 through P5, all recruited from customers who use the mobile scanner daily. The question was how they recover from a failed scan. During the task, P1 and P2 both restarted the app before reading the message because they assumed every scan error was temporary. P3 read the message and corrected the item code. P4 said that she usually asks a colleague for help, but we did not observe her recovery process. P5 completed the task without a scan error. The researcher thought the error wording might contribute to restarting, but this was not tested. We heard several stories about busy shifts, though none established whether workload changed recovery behavior. We did not interview occasional scanner users.

After 78 words

Sample: Five daily-use warehouse supervisors (P1–P5); no occasional users. Findings - P1 and P2 restarted before reading the error; both assumed scan errors were temporary. - P3 read the message and corrected the item code. Exceptions - P4 reported asking a colleague; recovery was not observed. - P5 had no scan error. Open questions - Does error wording affect restarting? This is an untested researcher interpretation. - Does workload affect recovery? Busy-shift stories did not establish a link.

Examples illustrate the method. They do not measure model output.

Files and sources

An original sho.rten.it skill. Source notes.