Looking for the Whole Curve: An Inconclusive Pre-Specified Test of Mindsponge Theory's Inverted-U Prediction with Per-Person Trust in Advisors

Mindsponge theory (MT) predicts that people weigh the content of a message most when they neither trust nor distrust its source, and less as trust or distrust grows (prediction P2, an inverted U). A companion test supported the trusted half of this curve. Here we test the whole curve on the same open dataset (Vodrahalli et al., 2022), using a different sample and a different measure of trust. All 881 participants were told the advice was 80% accurate, so trust varied only through each person’s own view of whether the advisor’s kind, an AI or other people, is better at the task. That view spread evenly from clear distrust to clear trust. On the 8,611 trials in which a participant’s own first answer opposed the advice, we estimated how strongly that first judgment governed acceptance of the advice in three trust bins. The rule, its admission simulation, and one declared revision were fixed before any final answer was read by trust, and the test was run once. The weights were 2.19 (low trust), 2.51 (middle), and 1.91 (high trust): highest in the middle, as MT predicts, but the middle exceeded the low bin with z=1.00 and the high bin with z=1.95, short of the threshold of 2. The verdict is inconclusive. Post-verdict analyses sharpen the picture. On a continuous measure of acceptance, the fall from the middle to high trust is clear (z=2.64), while the middle and low bins are nearly equal (z=0.33). The data are about five times more probable under MT’s predicted fall on the trusted side than under no fall, and about four times more probable under no fall than under MT’s full-size fall on the distrust side. The trusted half of P2 is thus replicated in a new sample; the distrust half is not visible in these data, where “distrust” means only a mild, comparative judgment. The result does not refute MT, and it shows what a decisive test of the distrust half would need.

Authors

Institutions

Publication Details

Journal
Knowledge Commons Works (Michigan State University)
Published
2026-11-18
DOI
https://doi.org/10.17613/607xt-ghs25
Primary Topic
Decision-Making and Behavioral Economics
Type
article
Field-Weighted Citation Impact
0.00
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
article

Looking for the Whole Curve: An Inconclusive Pre-Specified Test of Mindsponge Theory's Inverted-U Prediction with Per-Person Trust in Advisors

Minh-Hoang Nguyen
5 links
Knowledge Commons Works (Michigan State University)
Decision-Making and Behavioral Economics
article

Looking for the Whole Curve: An Inconclusive Pre-Specified Test of Mindsponge Theory's Inverted-U Prediction with Per-Person Trust in Advisors

Minh-Hoang Nguyen
article en

Abstract

Mindsponge theory (MT) predicts that people weigh the content of a message most when they neither trust nor distrust its source, and less as trust or distrust grows (prediction P2, an inverted U). A companion test supported the trusted half of this curve. Here we test the whole curve on the same open dataset (Vodrahalli et al., 2022), using a different sample and a different measure of trust. All 881 participants were told the advice was 80% accurate, so trust varied only through each person’s own view of whether the advisor’s kind, an AI or other people, is better at the task. That view spread evenly from clear distrust to clear trust. On the 8,611 trials in which a participant’s own first answer opposed the advice, we estimated how strongly that first judgment governed acceptance of the advice in three trust bins. The rule, its admission simulation, and one declared revision were fixed before any final answer was read by trust, and the test was run once. The weights were 2.19 (low trust), 2.51 (middle), and 1.91 (high trust): highest in the middle, as MT predicts, but the middle exceeded the low bin with z=1.00 and the high bin with z=1.95, short of the threshold of 2. The verdict is inconclusive. Post-verdict analyses sharpen the picture. On a continuous measure of acceptance, the fall from the middle to high trust is clear (z=2.64), while the middle and low bins are nearly equal (z=0.33). The data are about five times more probable under MT’s predicted fall on the trusted side than under no fall, and about four times more probable under no fall than under MT’s full-size fall on the distrust side. The trusted half of P2 is thus replicated in a new sample; the distrust half is not visible in these data, where “distrust” means only a mild, comparative judgment. The result does not refute MT, and it shows what a decisive test of the distrust half would need.

Knowledge Commons Works (Michigan State University)
Phenikaa University (VN)
Decision-Making and Behavioral Economics
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon