Analyzing establishment nonresponse using an interpretable regression tree model with linked administrative data

DOI10.1214/11-AOAS521MaRDI QIDQ439171zbMATH OpenOpenAlexFDO

Authors Polly Phipps, Daniell Toth

Publication date 1 August 2012

Published in The Annals of Applied Statistics (Search for Journal in Brave)

Full work available at URL https://arxiv.org/abs/1206.6666, https://projecteuclid.org/euclid.aoas/1339419616

zbMATH Keywords

classification and regression trees (CART)establishment surveynonignorable nonresponsepropensity modelrecursive partitioning

Mathematics Subject Classification ID

Graphical methods in statistics (62A09) Sampling theory, sample surveys (62D05) Generalized linear models (logistic models) (62J12) Applications of statistics to economics (62P20)

Abstract: To gain insight into how characteristics of an establishment are associated with nonresponse, a recursive partitioning algorithm is applied to the Occupational Employment Statistics May 2006 survey data to build a regression tree. The tree models an establishment's propensity to respond to the survey given certain establishment characteristics. It provides mutually exclusive cells based on the characteristics with homogeneous response propensities. This makes it easy to identify interpretable associations between the characteristic variables and an establishment's propensity to respond, something not easily done using a logistic regression propensity model. We test the model obtained using the May data against data from the November 2006 Occupational Employment Statistics survey. Testing the model on a disjoint set of establishment data with a very large sample size $(n=179,360)$ offers evidence that the regression tree model accurately describes the association between the establishment characteristics and the response propensity for the OES survey. The accuracy of this modeling approach is compared to that of logistic regression through simulation. This representation is then used along with frame-level administrative wage data linked to sample data to investigate the possibility of nonresponse bias. We show that without proper adjustments the nonresponse does pose a risk of bias and is possibly nonignorable.

Recommendations

• Adjusting for nonresponse bias using logistic regression
• Regression adjustments for nonresponse
• Nonresponse bias adjustment in regression analysis
• Estimation of a cluster-level regression model under nonresponse within clusters
• Building consistent regression trees from complex sample data

Cites work

• Asymptotically efficient solutions to the classification problem
• Building consistent regression trees from complex sample data
• Comments on "Stability margin evaluation for uncertain linear systems
• Consistent nonparametric regression from recursive partitioning schemes
• scientific article; zbMATH DE number 3860199 (Why is no real title available?)
• Linear Model Selection by Cross-Validation
• Models for Nonresponse in Sample Surveys
• Nonresponse weighting adjustment using estimated response probability
• Selecting the amount of smoothing in nonparametric regression estimation for complex surveys
• Survey Nonresponse Adjustments for Estimates of Means
• The central role of the propensity score in observational studies for causal effects
• The elements of statistical learning. Data mining, inference, and prediction
• Using calibration weighting to adjust for nonignorable unit nonresponse

Cited in

(7)

• Variable selection in propensity score adjustment to mitigate selection bias in online surveys
• Adjusting models of ordered multinomial outcomes for nonignorable nonresponse in the occupational employment statistics survey
• Using the Fraction of Missing Information to Identify Auxiliary Variables for Imputation Procedures via Proxy Pattern‐mixture Models
• Analysing establishment survey non-response using administrative data and machine learning
• Statistical inference in the presence of imputed survey data through regression trees and random forests
• A review of tree-based methods for analyzing survey data
• Introducing web in a mixed-mode establishment survey: effects on nonresponse

Describes a project that uses

Uses Software

• ElemStatLearn

This page was built for publication: Analyzing establishment nonresponse using an interpretable regression tree model with linked administrative data

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q439171)