LLM-personalized weekly nudges cut electricity use by 0.56 kWh per room-day (18.3 percentage points more than text-only feedback) in a five-week randomized trial.
A large-scale replication of scenario-based experiments in psychology and management using large language models.Nature Computational Science, 5(8):627–634
2 Pith papers cite this work. Polarity classification is still indexing.
2
Pith papers citing it
citation-role summary
background 2
citation-polarity summary
fields
cs.CY 2years
2026 2verdicts
CONDITIONAL 2roles
background 2representative citing papers
Descriptive fit and causal fidelity diverge in LLM behavioral simulation: matching human response patterns does not ensure accurate treatment-effect estimates.
citing papers explorer
-
LLM-generated personalized nudges for improving pro-environmental behavior: Field evidence from resource conservation
LLM-personalized weekly nudges cut electricity use by 0.56 kWh per room-day (18.3 percentage points more than text-only feedback) in a five-week randomized trial.
-
Statistical realism is not evidence that LLMs can estimate treatment effects in social science experiments
Descriptive fit and causal fidelity diverge in LLM behavioral simulation: matching human response patterns does not ensure accurate treatment-effect estimates.