Aligning NLP Models with Target Population Perspectives using PAIR: Population-Aligned Instance Replication

TL;DR

This paper introduces PAIR, a post-processing method that adjusts training data to better reflect the target population's perspectives, improving model calibration without needing extra annotations.

Contribution

The paper presents PAIR, a novel data replication technique that enhances population representativity in NLP training data without additional annotation collection.

Findings

Non-representative annotator pools harm model calibration

PAIR effectively improves calibration by replicating underrepresented group annotations

Accuracy remains largely unaffected by the replication process

Abstract

Models trained on crowdsourced annotations may not reflect population views, if those who work as annotators do not represent the broader population. In this paper, we propose PAIR: Population-Aligned Instance Replication, a post-processing method that adjusts training data to better reflect target population characteristics without collecting additional annotations. Using simulation studies on offensive language and hate speech detection with varying annotator compositions, we show that non-representative pools degrade model calibration while leaving accuracy largely unchanged. PAIR corrects these calibration problems by replicating annotations from underrepresented annotator groups to match population proportions. We conclude with recommendations for improving the representativity of training data and model performance.