Treatment Effect

Predictive tool for social science experiments using LLMs.

Treatment Effect is a demo tool that utilizes Large Language Models (LLMs) to predict experimental treatment effects on survey outcomes for U.S. adults. Users can select dependent variables and input treatment messages, allowing them to simulate responses in a randomized controlled trial (RCT) setup. The application is designed to aid in the design of interventions and pilot testing of study materials, while also aiming to minimize harm to human participants by providing simulated predictions rather than requiring real human involvement. Its accuracy in predicting outcomes is based on assessments correlating closely with actual experimental data, making it a useful companion for researchers in the field.

Use case

This app helps researchers predict the effects of various treatments in social science experiments by simulating participant responses.

https://treatmenteffect.app

{{ ui.publicListPage.error === 'not_found' ? 'This collection is unavailable' : 'Failed to load' }}

{{ t('note.public') }}
{{ t('note.public_hint') }}

Название
Описание

Доступ

{{ t('list.public') }}
{{ t('list.public_hint') }}
{{ t('list.hide_owner') }}
{{ t('list.hide_owner_hint') }}

{{tag.name}}

{{ search.results.semantic ? t('cat.results') : t('cat.suggested') }}

{{app.title}}

{{app.subtitle}}

Treatment Effect

Use case

{{ ui.publicListPage.meta && ui.publicListPage.meta.name }}

{{ t('home_settings.title') }}

{{ t('home.apps') }} search

{{ pl.name }}

{{ t('search.title') }}

{{ t('home.collections') }}

{{ t('profile.title') }}

Новый лист

Treatment Effect

Use case

{{ ui.publicListPage.meta && ui.publicListPage.meta.name }}

{{ t('home.apps') }} search

{{ pl.name }}

{{ t('search.title') }}

{{ t('home.collections') }}

{{ t('profile.title') }} gear person_crop_circle {{ t('settings.sign_in_hint') }}

Новый лист

{{ t('profile.title') }}