# Explaining Paranoia: Computational Modeling of Belief Updating in the General Population

> **NIH NIH R21** · YALE UNIVERSITY · 2020 · $183,750

## Abstract

!
Paranoia, the unfounded belief that others have bad intentions towards us, can be a very distressing feature of
schizophreniform psychotic illnesses. They are trypically with D2 dopamine receptor blocking drugs. However
up to 30% of patients may experiences residual hallucinations despite adequate dosing and those that do are
at higher risk of harm to themselves and others. We propose to address this unmet clinical need through
computational psychiatry. We have devised a computer-based ask that assays belief updating. When they are
uncertain about the best choice, individuals with paranoia fail to update their beliefs appropriatel.. Our
computational analyses of participant behavior reveal that people who are paranoid are resistant to belief
updating because of lower learning rates. We seek to repolicate and extend these preliminary observations in
a large online sample. Three specific aims are proposed: Specific Aim 1. Acquire a large (n=800) behavioral
sample of reversal learning performance in high (n=400) and low (n=400) paranoia sample. 200 subjects will
perform each of four different reversal learning tasks differing in their underlying reward contingencies, whether
and how those contingencies chgange across task blocks. We predict a specific association between high
paranoia and poor performance on the task version in which uncertainty increases unexpectedly. Specific Aim
2. We will fit a hierarchical computational model to participant behavior and estimate their learning rates to
parameterize their belief updating, comparing these parameters in high and low paranoia individuals.,Specific
Aim 3. We will fit alternate models (from preclinical animal work on reversal learning) and compare model fits
to those achieved with the hierarchical model in Aim 2. We expect hierarchical belief updating deficits will be
pronounced in paranoid individuals. If we are correct, the data that we gather will be the first step towards task
and computational model-based development of new treatments for paranoia, as well as diversion of patients
towards those treatmenst based on computational parameters. However, even if we are wrong, we will learn
that the belief updating model is lacking and that we should pursue other explanations for paranoia.
 !

## Key facts

- **NIH application ID:** 9985203
- **Project number:** 5R21MH120799-02
- **Recipient organization:** YALE UNIVERSITY
- **Principal Investigator:** PHILIP Robert CORLETT
- **Activity code:** R21 (R01, R21, SBIR, etc.)
- **Funding institute:** NIH
- **Fiscal year:** 2020
- **Award amount:** $183,750
- **Award type:** 5
- **Project period:** 2019-08-01 → 2021-11-30

## Primary source

NIH RePORTER: https://reporter.nih.gov/project-details/9985203

## Citation

> US National Institutes of Health, RePORTER application 9985203, Explaining Paranoia: Computational Modeling of Belief Updating in the General Population (5R21MH120799-02). Retrieved via AI Analytics 2026-05-24 from https://api.ai-analytics.org/grant/nih/9985203. Licensed CC0.

---

*[NIH grants dataset](/datasets/nih-grants) · CC0 1.0*