A stableness of resistance model for nonresponse adjustment with callback data

We propose a stableness of resistance assumption for nonresponse adjustment with callback data—a traditional form of paradata that are available in almost all modern surveys to track the data collection process. We establish the identification and the semiparametric efficiency theory without imposing any parametric restrictions, and propose a suite of semiparametric estimation methods including doubly robust ones, which generalize existing parametric approaches for using callback data. We also consider extension of this framework to causal inference for unmeasured confounding adjustment. Application to a Consumer Expenditure Survey dataset suggests an association between nonresponse and high housing expenditures, and reanalysis of Card (1995)’s dataset on the return to schooling shows a smaller effect of education in the overall population than in the respondents.

[1]  Stijn Vansteelandt,et al.  Bias-Reduced Doubly Robust Estimation , 2015 .

[2]  Mick P. Couper,et al.  MEASURING SURVEY QUALITY IN A CASIC ENVIRONMENT , 2002 .

[3]  Stijn Vansteelandt,et al.  Inference for treatment effect parameters in potentially misspecified high-dimensional models , 2020 .

[4]  G. Imbens,et al.  Approximate residual balancing: debiased inference of average treatment effects in high dimensions , 2016, 1604.07125.

[5]  J. Qin,et al.  Semiparametric maximum likelihood inference by using failed contact attempts to adjust for nonignorable nonresponse , 2014 .

[6]  Z. Geng,et al.  Identifying Causal Effects With Proxy Variables of an Unmeasured Confounder. , 2016, Biometrika.

[7]  Ilya Shpitser,et al.  Semiparametric Inference for Non-monotone Missing-Not-at-Random Data: the No Self-Censoring Model , 2019 .

[8]  Eric J Tchetgen Tchetgen,et al.  A general instrumental variable framework for regression analysis with outcome missing not at random , 2017, Biometrics.

[9]  Joshua D. Angrist,et al.  Identification of Causal Effects Using Instrumental Variables , 1993 .

[10]  Zhiqiang Tan,et al.  A Distributional Approach for Causal Inference Using Propensity Scores , 2006 .

[11]  Wei Feng,et al.  Pattern mixture models for the analysis of repeated attempt designs , 2015, Biometrics.

[12]  N. C. Schaeffer,et al.  Institute for Research on Poverty Discussion Paper no. 1024-93 Using Survey Participants to Estimate the Impact of Nonparticipation , 1993 .

[13]  Jun Shao,et al.  Semiparametric Pseudo-Likelihoods in Generalized Linear Models With Nonignorable Missing Data , 2015 .

[14]  Alisa J. Stephens,et al.  Locally Efficient Estimation of Marginal Treatment Effects When Outcomes Are Correlated: Is the Prize Worth the Chase? , 2014, The international journal of biostatistics.

[15]  Michael Peress.  Correcting for Survey Nonresponse Using Variable Response Propensity , 2010 .

[16]  Juha M. Alho,et al.  Adjusting for nonresponse bias using logistic regression , 1990 .

[17]  Marie Davidian,et al.  Improved Doubly Robust Estimation When Data Are Monotonely Coarsened, with Application to Longitudinal Studies with Dropout , 2011, Biometrics.

[18]  J. Robins,et al.  IDENTIFICATION AND INFERENCE FOR MARGINAL AVERAGE TREATMENT EFFECT ON THE TREATED WITH AN INSTRUMENTAL VARIABLE. , 2015, Statistica Sinica.

[19]  F. Filion,et al.  Exploring and Correcting for Nonresponse Bias Using Follow-ups of Non Respondents , 1976 .

[20]  James M. Robins,et al.  Characterization of parameters with a mixed bias property , 2019, Biometrika.

[21]  Judea Pearl,et al.  Graphical Models for Processing Missing Data , 2018, Journal of the American Statistical Association.

[22]  Eric Tchetgen Tchetgen,et al.  Bounded, efficient and multiply robust estimation of average treatment effects using instrumental variables , 2016, Journal of the Royal Statistical Society. Series B, Statistical methodology.

[23]  Revisiting the continuum of resistance model in the digital age: A comparison of early and delayed respondents to the Norwegian Counties Public Health Survey , 2020 .

[24]  Elizabeth L. Ogburn,et al.  Doubly robust estimation of the local average treatment effect curve , 2015, Journal of the Royal Statistical Society. Series B, Statistical methodology.

[25]  M. Kenward,et al.  Every missingness not at random model has a missingness at random counterpart with equal fit , 2008 .

[26]  W. Newey,et al.  Large sample estimation and hypothesis testing , 1986 .

[27]  Dan Jackson,et al.  How much can we learn about missing data?: an exploration of a clinical trial in psychiatry , 2010, Journal of the Royal Statistical Society. Series A,.

[28]  A. Winsor.  Sampling techniques. , 2000, Nursing times.

[29]  Matthew Hotopf,et al.  Using number of failed contact attempts to adjust for non‐ignorable non‐response , 2006 .

[30]  Jerome P. Reiter,et al.  Itemwise conditionally independent nonresponse modeling for incomplete multivariate data , 2016, 1609.00656.

[31]  Jae Kwang Kim,et al.  Propensity score adjustment with several follow-ups , 2014 .

[32]  M. J. van der Laan,et al.  The International Journal of Biostatistics Targeted Maximum Likelihood Learning , 2011 .

[33]  Eric J Tchetgen Tchetgen,et al.  Semiparametric Estimation with Data Missing Not at Random Using an Instrumental Variable. , 2016, Statistica Sinica.

[34]  Jae Kwang Kim,et al.  A Semiparametric Estimation of Mean Functionals With Nonignorable Missing Data , 2011 .

[35]  L. Hansen.  Large Sample Properties of Generalized Method of Moments Estimators , 1982 .

[36]  Joseph G. Ibrahim,et al.  A Weighted Estimating Equation for Missing Covariate Data with Properties Similar to Maximum Likelihood , 1999 .

[37]  K. Olson,et al.  Paradata for Nonresponse Adjustment , 2013 .

[38]  Zhiqiang Tan,et al.  Model-assisted inference for treatment effects using regularized calibrated estimation with high-dimensional data , 2018, The Annals of Statistics.

[39]  Alexander D'Amour,et al.  Flexible Sensitivity Analysis for Observational Studies Without Observable Implications , 2018, Journal of the American Statistical Association.

[40]  Xavier D'Haultfoeuille,et al.  A New Instrumental Method for Dealing with Endogenous Selection , 2010 .

[41]  J. Robins,et al.  Doubly Robust Estimation in Missing Data and Causal Inference Models , 2005, Biometrics.

[42]  Hua Yun Chen.  A Semiparametric Odds Ratio Model for Measuring Association , 2007, Biometrics.

[43]  G. Osius.  The association between two random elements: A complete characterization and odds ratio models , 2004 .

[44]  J. Heckman.  Sample selection bias as a specification error , 1979 .

[45]  E. Ziegel,et al.  Nonresponse In Household Interview Surveys , 1998 .

[46]  Kenneth G. Manton,et al.  Correcting for nonavailability bias in surveys by weighting based on number of callbacks , 1993 .

[47]  Wang Miao,et al.  On varieties of doubly robust estimators under missingness not at random with a shadow variable , 2015, Biometrika.

[48]  James M. Robins,et al.  DOUBLY ROBUST INSTRUMENTAL VARIABLE REGRESSION , 2012 .

[49]  Henian Chen,et al.  BAYESIAN INFERENCE FOR NONRESPONSE TWO-PHASE SAMPLING , 2018 .

[50]  Edward H Kennedy,et al.  Non‐parametric methods for doubly robust estimation of continuous treatment effects , 2015, Journal of the Royal Statistical Society. Series B, Statistical methodology.

[51]  W. Deming.  On a Probability Mechanism to Attain an Economic Balance Between the Resultant Error of Response and the Bias of Nonresponse , 1953 .

[52]  A. Politz,et al.  An Attempt to Get the “Not at Homes” into the Sample Without Callbacks , 1949 .

[53]  F. Kreuter.  Improving Surveys with Paradata , 2013 .

[54]  David Card,et al.  Using Geographic Variation in College Proximity to Estimate the Return to Schooling , 1993 .

[55]  J. Robins,et al.  Sensitivity Analysis for Selection bias and unmeasured Confounding in missing Data and Causal inference models , 2000 .

[56]  Eric J. Tchetgen Tchetgen,et al.  Multiply robust causal inference with double‐negative control adjustment for categorical unmeasured confounding , 2018, Journal of the Royal Statistical Society. Series B, Statistical methodology.

[57]  Zhiqiang Tan,et al.  Bounded, efficient and doubly robust estimation with inverse weighting , 2010 .

[58]  Roderick J. A. Little,et al.  Subsampling Callbacks to Improve Survey Efficiency , 2000 .

[59]  Paul P. Biemer,et al.  Using level‐of‐effort paradata in non‐response adjustments with application to field surveys , 2013 .

[60]  M. J. Laan,et al.  Doubly robust nonparametric inference on the average treatment effect , 2017, Biometrika.

[61]  Joseph Kang,et al.  Demystifying Double Robustness: A Comparison of Alternative Strategies for Estimating a Population Mean from Incomplete Data , 2007, 0804.2958.

[62]  E. J. Tchetgen Tchetgen,et al.  An Introduction to Proximal Causal Learning , 2020, medRxiv.

[63]  Zhi Geng,et al.  Identifiability of Normal and Normal Mixture Models with Nonignorable Missing Data , 2015, 1509.03860.

[64]  J. Qin,et al.  Generalization of Heckman selection model to nonignorable nonresponse using call-back information , 2018 .

[65]  J. Qin,et al.  Semiparametric maximum likelihood inference for nonignorable nonresponse with callbacks , 2018 .

[66]  J. Robins,et al.  Adjusting for Nonignorable Drop-Out Using Semiparametric Nonresponse Models , 1999 .

[67]  Jae Kwang Kim,et al.  An Instrumental Variable Approach for Identification and Estimation with Nonignorable Nonresponse , 2014 .

[68]  J. Connor,et al.  Assessment of Non-Response Bias in Estimates of Alcohol Consumption: Applying the Continuum of Resistance Model in a General Population Survey in England , 2017, PloS one.

[69]  Andrea Rotnitzky,et al.  Estimation of regression models for the mean of repeated outcomes under nonignorable nonmonotone nonresponse. , 2007, Biometrika.

[70]  D. Rubin.  Estimating causal effects of treatments in randomized and nonrandomized studies. , 1974 .