CMStatistics 2022: Start Registration
View Submission - CMStatistics
B0419
Title: The penalized instrumental variables methods for many invalid instruments Authors:  Muhammad Qasim - Jonkoping University (Sweden) [presenting]
Kristofer Mansson - JU JIBS (Sweden)
Abstract: The valid instrumental variables must not directly affect the outcome variable and not be correlated to unmeasured variables. But practically, instrumental variables (IV) are likely to be invalid. We derive a LASSO procedure for the k-class IV estimation methods in the linear IV model. The proposed method is robust for estimating the causal effect in the presence of many/weak invalid and valid instruments, with theoretical assurances on its execution. In addition, a two-step numerical algorithm is developed to estimate causal effects. The performance of the proposed penalized k-class IV methods is assessed by Monte Carlo simulation.