More power to you: Using machine learning to augment human coding for more efficient inference in text-based randomized trials

TL;DR

This paper introduces a machine learning-based framework that enhances the power of impact assessments in text-based randomized trials by leveraging untapped unscored documents, reducing human coding efforts while maintaining statistical power.

Contribution

It presents a novel inferential framework combining causal inference, survey sampling, and machine learning to improve impact estimation efficiency in text-based trials.

Findings

Reduces human coding effort while maintaining power.

Effective in simulation and real-world education trial.

Provides unbiased treatment impact estimates.

Abstract

For randomized trials that use text as an outcome, traditional approaches for assessing treatment impact require that each document first be manually coded for constructs of interest by trained human raters. This process, the current standard, is both time-consuming and limiting: even the largest human coding efforts are typically constrained to measure only a small set of dimensions across a subsample of available texts. In this work, we present an inferential framework that can be used to increase the power of an impact assessment, given a fixed human-coding budget, by taking advantage of any "untapped" observations -- those documents not manually scored due to time or resource constraints -- as a supplementary resource. Our approach, a methodological combination of causal inference, survey sampling methods, and machine learning, has four steps: (1) select and code a sample of documents; (2) build a machine learning model to predict the human-coded outcomes from a set of automatically extracted text features; (3) generate machine-predicted scores for all documents and use these scores to estimate treatment impacts; and (4) adjust the final impact estimates using the residual differences between human-coded and machine-predicted outcomes. This final step ensures any biases in the modeling procedure do not propagate to biases in final estimated effects. Through an extensive simulation study and an application to a recent field trial in education, we show that our proposed approach can be used to reduce the scope of a human-coding effort while maintaining nominal power to detect a significant treatment impact.