Leaning on Habit Under Pressure: An Inconclusive but Informative Test of Whether Mindsponge Theory's Trust Route Scales with Habit Strength
Description
Mindsponge theory (MT), in the formal reading of its testing program, implies that when evaluation becomes costly, people lean toward what they already trust, and that the size of this lean is proportional to how strongly they trust it. We tested this proportionality on open data in which 73 people made repeated lottery choices under three levels of time pressure after a priming phase that built a habit toward either the safe or the risky option. Habit strength, measured in the priming phase, served as the trust advantage of the habitual option. The rule, an admission simulation, and one declared revision were fixed before choices were analyzed by habit strength. The estimated scaling was 0.159 (SE 0.130), 82% of the value implied by MT, with a 95% confidence interval from −0.096 to 0.414. The verdict under the pre-specified rule is inconclusive. The result nonetheless leans toward MT. In a breakdown by thirds of habit strength, the rise in lean under high pressure grew from 0.14 to 0.49 to 0.70, close to the line MT predicts. The data favor MT over an account with a separate habit channel only weakly (likelihood ratio about 2), but favor it over a pure loss of precision strongly (likelihood ratio about 470). An inconclusive verdict is not a refutation: the estimate sits near MT’s prediction, and the test was too small to tell it apart from no scaling. We estimate that about 260 to 380 participants would be needed for a decisive test, and we argue that such a test is worth running because it bears directly on what MT’s trust route is.