How often do our two-group decision rules report a false gap when the groups' extreme values differ?
Study 86 · No reading (stopped by its own checks) · recorded 4 November 2026 · answer: Not settled · full paper with figures
The question
Earlier studies chose between two groups by rules pairing the plain average with the clipped one (which pulls each group's extreme values in) and one rule widened its ranges; the clipped average alone was shown to report false gaps when one group's extremes reach further. We ask how often each whole rule, applied step by step, reports a false gap.
Rules fixed before the result
Fixed before any test: on made-up comparisons with no true difference, a rule failed when its rate of false gaps was shown above 7.5%, or above 3.75% on either side, and held when its 95% range stayed at most 7.5% and 3.75% on each side. Each rule was to be read on 6,000 comparisons at its own groups. Before any rate, study 67's rule had to return its record's 68 false gaps in 2,000, any difference stopping the study.
The answer
Not settled: the study stopped at a check planned before any rate of false gaps was measured, after the rebuilt days had passed their identity checks. The plan required that study 67's rule, rebuilt here on the first 2,000 made-up comparisons with no true difference and with both groups' extreme values reaching equally far, report a false gap on exactly the 68 that study's own record counts; it reported a different number. The plan fixed how far each range is widened at the values measured on that study's events, while its record set the widening afresh on each comparison, so the two counts could not agree. By the rule set in advance, any difference stopped the study. Nothing is shown about how often any of the rules reports a false gap, in either direction.
What it changes
No rule and no published answer changed. Under the rule set in advance for a stop, we decided that a corrected design follows as a later study: it widens the ranges on each comparison as study 67's own record did, so that the first check compares like with like.
What remains uncertain
No rate of false gaps was measured, so the question stays open for every rule studied. These days were studied before, so even a measured rate would have been a check within our development data, not a test on fresh data. The rates would describe made-up comparisons on ordinary days, not the earlier studies' event days.