Causal Effects of Linguistic Properties

TL;DR

This paper introduces TextCause, a novel method for estimating the causal effects of linguistic properties from observational data, addressing challenges of noisy proxies and text adjustment, with applications in review sentiment and bureaucratic response times.

Contribution

It formalizes the causal effect of linguistic properties, proposes a bias-bounded estimator using language models and distant supervision, and introduces the TextCause algorithm.

Findings

TextCause outperforms related methods in semi-simulated experiments.

The method effectively estimates the impact of review sentiment on sales.

Applied case study shows politeness affects bureaucratic response times.

Abstract

We consider the problem of using observational data to estimate the causal effects of linguistic properties. For example, does writing a complaint politely lead to a faster response time? How much will a positive product review increase sales? This paper addresses two technical challenges related to the problem before developing a practical method. First, we formalize the causal quantity of interest as the effect of a writer's intent, and establish the assumptions necessary to identify this from observational data. Second, in practice, we only have access to noisy proxies for the linguistic properties of interest -- e.g., predictions from classifiers and lexicons. We propose an estimator for this setting and prove that its bias is bounded when we perform an adjustment for the text. Based on these results, we introduce TextCause, an algorithm for estimating causal effects of linguistic properties. The method leverages (1) distant supervision to improve the quality of noisy proxies, and (2) a pre-trained language model (BERT) to adjust for the text. We show that the proposed method outperforms related approaches when estimating the effect of Amazon review sentiment on semi-simulated sales figures. Finally, we present an applied case study investigating the effects of complaint politeness on bureaucratic response times.