Work

Identifying Commitment-Gated Choice Reversal:Exact Limits and Design Requirements

Year: 2026
Type: preprint
Abstract: When does prior commitment make a choice genuinely harder to reverse rather than simply less attractive? This article develops Structural Revisability Theory (SRT) as a restricted process account of history-dependent choice reversal and then asks whether its distinctive commitment gate is identifiable. Three results follow. First, an exact identification limit: in the reduced gated model U = α + A[1 − exp{−ρdq(1 − K 0 )}], fixing initial accessibility K 0 makes the reparameterization ρ′ = ρ(1 − K 0 ) reproduce a generic saturating rival at every dwell and activation level. No sample size breaks that equivalence. Second, in an equal-complexity synthetic binary-choice comparison, merely crossing two accessibility levels yields weak discrimination: across 800 synthetic datasets, pooled correct selection is 63.8% (95% CI 60.4–67.0). Third, a one-factor-at-a-time decomposition identifies the useful design variation. Widening initial accessibility alone raises recovery to 83.9% (81.2–86.3), whereas longer dwell adds only 0.3 percentage points. The result is methodological and theoretical rather than behavioral evidence: tests of commitment-gated reversal must vary independently measured initial reversibility, not simply observe persistence for longer.
Author: Daniel Bryan
Language: English
Open Access status: green
OpenAlex ID: W7226013022
Cites: 23
Cited by:
Related to: 10