Causal Inference in Repeated Observational Studies: A Case Study of eBay Product Releases

TL;DR

This paper presents a Bayesian framework for causal inference in observational studies with sequential treatments, demonstrated through eBay product release data, addressing confounding factors like early-adopter effects.

Contribution

The authors introduce a flexible hierarchical Bayesian method to control for early-adopter effects in sequential treatment analysis, improving causal estimate accuracy.

Findings

Naive causal estimates are highly misleading.

The proposed method provides stable, sensible causal estimates.

The approach performs well in out-of-sample validation.

Abstract

Causal inference in observational studies is notoriously difficult, due to the fact that the experimenter is not in charge of the treatment assignment mechanism. Many potential con- founding factors (PCFs) exist in such a scenario, and if one seeks to estimate the causal effect of the treatment on a response, one needs to control for such factors. Identifying all relevant PCFs may be difficult (or impossible) given a single observational study. Instead, we argue that if one can observe a sequence of similar treatments over the course of a lengthy time period, one can identify patterns of behavior in the experimental subjects that are correlated with the response of interest and control for those patterns directly. Specifically, in our case-study we find and control for an early-adopter effect: the scenario in which the magnitude of the response is highly correlated with how quickly one adopts a treatment after its release. We provide a flexible hierarchical Bayesian framework that controls for such early-adopter effects in the analysis of the effects of multiple sequential treatments. The methods are presented and evaluated in the context of a detailed case-study involving product updates (newer versions of the same product) from eBay, Inc. The users in our study upgrade (or not) to a new version of the product at their own volition and timing. Our response variable is a measure of user actions, and we study the behavior of a large set of users (n = 10.5 million) in a targeted subset of eBay categories over a period of one year. We find that (a) naive causal estimates are hugely misleading and (b) our method, which is relatively insensitive to modeling assumptions and exhibits good out-of-sample predictive validation, yields sensible causal estimates that offer eBay a stable basis for decision-making.