The sample mean is 6 minutes, while the median is 3.5. Excluding T6 changes the mean to 3.6; that is a sensitivity result, not a cleaned truth. The next useful action is broader, consistently collected evidence rather than an immediate causal verdict.

Critical check: No row is silently removed and no sample result is presented as proven causation or a population guarantee.

Final project: Deliver the raw table, summaries, dot plot, outlier log and decision memo. Add waits of 5 and 7 from the next shift, recalculate all summaries and explain why the expanded convenience sample still has limits.