How Does Ego-Relevance Affect Nuisance Neglect? The Role of Motivated Skepticism

How to cite

Wang, Shuyu; Ye, Maoliang. 2026. "How Does Ego-Relevance Affect Nuisance Neglect? The Role of Motivated Skepticism." AEA RCT Registry. AEARCTR-0017553. https://doi.org/10.1257/rct.17553-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Shuyu Wang* , Maoliang Ye

* Primary investigator

Registry ID
AEARCTR-0017553
DOI
https://doi.org/10.1257/rct.17553-1.0
First registered
Last updated
Status
In development
Trial dates
2026-09-07 – 2027-08-01
Countries
Private
JEL codes
C91, D83, D91
Registry record
http://www.socialscienceregistry.org/trials/17553

Abstract

Do people account for payoff-irrelevant but signal-relevant variables (i.e., nuisance variables) when interpreting information, and does ego-relevance influence this inference process?

We study nuisance neglect in a transparent two-component signal-generating structure in which a payoff-relevant rank state (X) and a payoff-irrelevant but signal-relevant nuisance variable (Y) jointly determine the observed signal (S).

In the experiment, each subject completes two parallel tasks in randomized order: an IQ-style task, which is ego-relevant, and a temperature-guessing task, which is ego-irrelevant. In each task, we elicit subjects’ prior beliefs about their rank state, provide a signal generated by a known rule that combines the true state and the nuisance variable, and then elicit posterior beliefs.

We randomize the posterior elicitation condition between subjects. In the Narrow condition, subjects report posterior beliefs about the state only. In the Broad condition, subjects first report which distributions of the nuisance variable are objectively compatible with the observed signal, and then report posterior beliefs about the state. This condition makes the nuisance variable salient and encourages subjects to consider how it enters the signal-generating process.

We measure nuisance neglect using the relative distance between subjects’ posterior beliefs and the Bayesian and fully nuisance-neglect benchmarks. Our main analysis tests whether ego-relevance affects nuisance neglect differently across signal valence. We hypothesize that ego-relevance increases nuisance neglect for good signals but decreases nuisance neglect for bad signals.

Intervention

First, ego relevance is varied through the task. Each subject completes two tasks in randomized order: an IQ-style task, which is ego-relevant, and a temperature-guessing task, which is ego-irrelevant. In each task, subjects report beliefs about their performance rank in that task. This structure allows us to compare belief updating across ego-relevant and ego-irrelevant settings while using all task-level observations. Task order is randomized to help address possible order effects.

Second, posterior elicitation condition is randomized between subjects through the elicitation format. In the Narrow condition, subjects report posterior beliefs about the rank state only. In the Broad condition, subjects first report which distributions of the nuisance variable are objectively compatible with the observed signal under the known signal-generating rule. They then report posterior beliefs about the rank state.

Experimental design

Each subject completes two task modules in randomized order: an IQ-style task module and a temperature-guessing task module.

In each module, subjects first complete the task and are then assigned to a newly formed four-person comparison group. Group assignment is conducted separately in the two modules. They then report a prior belief distribution over their own performance rank within that group. There are four possible performance ranks. Next, subjects observe a rank-related signal. The signal is determined jointly by the subject’s realized rank and a payoff-irrelevant but signal-relevant nuisance variable. Subjects then report beliefs using one of two posterior elicitation conditions.

Each subject is randomly assigned to one posterior elicitation condition, either Narrow or Broad, and uses that same condition in both task modules. In the Narrow condition, subjects report posterior beliefs about their performance rank only. In the Broad condition, subjects report which distributions of the nuisance variable are objectively compatible with the observed signal under the known signal-generating rule, in addition to reporting posterior beliefs about their performance rank.

At the end of the second module, subjects are given an opportunity to purchase an additional ranking signal generated using an independent draw of the nuisance variable and, if they obtain it, submit an additional posterior belief report.

At the end of the experiment, subjects complete a short questionnaire collecting demographics and background characteristics, task-specific intensity of ego relevance, and self-reported consideration of the nuisance variable.

Design details

Randomization method
Computer-generated randomization implemented by the experimental software.
Randomization unit
Individual.
Clusters
N/A
Observations
About 400 subjects (university students).
Arms
We plan to recruit 400 subjects in total. Subjects are assigned to one of four experimental cells defined by posterior elicitation condition and task order: Narrow condition, IQ task first, temperature-guessing task second: approximately 152 subjects. Narrow condition, temperature-guessing task first, IQ task second: approximately 152 subjects. Broad condition, IQ task first, temperature-guessing task second: approximately 48 subjects. Broad condition, temperature-guessing task first, IQ task second: approximately 48 subjects. The allocation prioritizes the Narrow condition, which is the focus of the primary analysis examining the effect of ego-relevance on nuisance neglect across signal valence. The Broad condition receives a smaller allocation and is used for the mechanism analysis examining whether making the nuisance variable salient reduces nuisance neglect.
Primary outcomes
Prior beliefs, posterior beliefs, compatible nuisance-variable distributions, nuisance neglect index
Secondary outcomes
Realized rank, gender, age, education level, major, statistics coursework exposure, intensity of ego relevance, self-reported consideration of the nuisance variable, response time in posterior belief elicitation, willingness to pay for an additional ranking signal and extra posterior belief after receiving the additional signal.

Cited by

No citation data available for this registration yet.

Keywords