TSCI: two stage curvature identification for causal inference with invalid instruments

TL;DR

TSCI introduces a two-stage method for causal inference that effectively estimates treatment effects even with invalid instruments by combining machine learning and data-adaptive violation correction.

Contribution

It provides a novel two-stage algorithm that relaxes classical instrumental variable assumptions, enabling valid causal inference with potentially invalid instruments.

Findings

Effective treatment effect estimation with all invalid instruments.

Utilizes machine learning to handle nonlinearities in the treatment model.

Data-adaptive selection of instrument violation space.

Abstract

TSCI implements treatment effect estimation from observational data under invalid instruments in the R statistical computing environment. Existing instrumental variable approaches rely on arguably strong and untestable identification assumptions, which limits their practical application. TSCI does not require the classical instrumental variable identification conditions and is effective even if all instruments are invalid. TSCI implements a two-stage algorithm. In the first stage, machine learning is used to cope with nonlinearities and interactions in the treatment model. In the second stage, a space to capture the instrument violations is selected in a data-adaptive way. These violations are then projected out to estimate the treatment effect.