arXiv subjects

Daniell Toth

Publications and source records attributed to Daniell Toth.

7 recordsLinked to original sources

arXiv2025

Statistics-Friendly Confidentiality Protection for Establishment Data, with Applications to the QCEW

Confidentiality for business data is an understudied area of disclosure avoidance, where legacy methods struggle to provide acceptable results. Standard formal privacy techniques for person-level data, like differential privacy, are designed to protect against membership inference and hence do not provide suitable confidentiality/utility trade-offs due to the highly skewed nature of business data and because extreme outlier records are often important contributors to query answers. Prior proposals, therefore, took a personalized differential privacy approach that allowed privacy parameters to degrade for the outlying records -- larger establishments get weaker membership inference guarantees. However, providing guarantees to some entities that are strictly weaker than guarantees for others is problematic from a policy standpoint. In this paper, we propose a novel confidentiality framework for business data with a focus on interpretability for policy makers. Instead of protecting against membership inference, which is often not a concern in business data, we protect against attribute inferences that are too precise. In our framework, data curators specify a neighbor function that is used to define uncertainty interval bands around an establishment's attribute values and the privacy parameters govern the strength of indistinguishability between values within the same uncertainty interval.We propose two query-answering mechanisms under this framework and evaluate them on: (1) a confidential Quarterly Census of Employment and Wages (QCEW) dataset produced by the U.S. Bureau of Labor Statistics (this was done through a cooperative agreement), and (2) a substitute dataset that we created from public sources (and will publicly release).

cs.CR↗

arXiv2017

Automated Selection of Post-Strata using a Model-Assisted Regression Tree Estimator

Auxiliary information can increase the efficiency of survey estimators through an assisting model when the model captures some of the relationship between the auxiliary data and the study variables. Despite their superior properties, model-assisted estimators are rarely used in anything but their simplest form by statistical agencies to produce official statistics. This is due to the fact that the more complicated models that have been used in model-assisted estimation are often ill suited to the available auxiliary data. Under a model-assisted framework, we propose a regression tree estimator for a finite population total. Regression tree models are adept at handling the type of auxiliary data usually available in the sampling frame and provide a model that is easy to explain and justify. The estimator can be viewed as a post-stratification estimator where the post-strata are automatically selected by the recursive partitioning algorithm of the regression tree. We establish consistency of the regression tree estimator and compare its performance to other survey estimators using the US Bureau of Labor Statistics Occupational Employment Statistics Survey.

stat.ME↗

arXiv2017

A Permutation Test on Complex Sample Data

Permutation tests are a distribution free way of performing hypothesis tests. These tests rely on the condition that the observed data are exchangeable among the groups being tested under the null hypothesis. This assumption is easily satisfied for data obtained from a simple random sample or a controlled study after simple adjustments to the data, but there is no general method for adjusting survey data collected using a complex sample design to allow for permutation tests. In this article, we propose a general method for performing a pseudo- permutation test that accounts for the complex sample design. The proposed method is not a true permutation test in that the new values do not come from the set of observed values in general, but of an expanded set of values satisfying a random-effects model on the clustered residuals. Tests using a simulated population comparing the performance of the proposed method to permutation tests that ignore the sample design demonstrate that it is necessary to account for certain design features in order to obtain reasonable p-value estimates.

stat.ME↗

arXiv2016

Bayesian Estimation Under Informative Sampling

Bayesian analysis is increasingly popular for use in social science and other application areas where the data are observations from an informative sample. An informative sampling design leads to inclusion probabilities that are correlated with the response variable of interest. Model inference performed on the observed sample taken from the population will be biased for the population generative model under informative sampling since the balance of information in the sample data is different from that for the population. Typical approaches to account for an informative sampling design under Bayesian estimation are often difficult to implement because they require re-parameterization of the hypothesized generating model, or focus on design, rather than model-based, inference. We propose to construct a pseudo-posterior distribution that utilizes sampling weights based on the marginal inclusion probabilities to exponentiate the likelihood contribution of each sampled unit, which weights the information in the sample back to the population. Our approach provides a nearly automated estimation procedure applicable to any model specified by the data analyst for the population and retains the population model parameterization and posterior sampling geometry. We construct conditions on known marginal and pairwise inclusion probabilities that define a class of sampling designs where $L_{1}$ consistency of the pseudo posterior is guaranteed. We demonstrate our method on an application concerning the Bureau of Labor Statistics Job Openings and Labor Turnover Survey.

math.ST↗

arXiv2014

Adjusting models of ordered multinomial outcomes for nonignorable nonresponse in the occupational employment statistics survey

An establishment's average wage, computed from administrative wage data, has been found to be related to occupational wages. These occupational wages are a primary outcome variable for the Bureau of Labor Statistics Occupational Employment Statistics survey. Motivated by the fact that nonresponse in this survey is associated with average wage even after accounting for other establishment characteristics, we propose a method that uses the administrative data for imputing missing occupational wage values due to nonresponse. This imputation is complicated by the structure of the data. Since occupational wage data is collected in the form of counts of employees in predefined wage ranges for each occupation, weighting approaches to deal with nonresponse do not adequately adjust the estimates for certain domains of estimation. To preserve the current data structure, we propose a method to impute each missing establishment's wage interval count data as an ordered multinomial random variable using a separate survival model for each occupation. Each model incorporates known auxiliary information for each establishment associated with the distribution of the occupational wage data, including geographic and industry characteristics. This flexible model allows the baseline hazard to vary by occupation while allowing predictors to adjust the probabilities of an employee's salary falling within the specified ranges. An empirical study and simulation results suggest that the method imputes missing OES wages that are associated with the average wage of the establishment in a way that more closely resembles the observed association.

stat.AP↗

arXiv2014

Building Consistent Regression Trees From Complex Sample Data

In the past several years a wide range of methods for the construction of regression trees and other estimators based on the recursive partitioning of samples have appeared in the statistics literature. Many applications involve data collected through a complex sample design. At present, however, relatively little is known regarding the properties of these methods under complex designs. This article proposes a method for incorporating information about the complex sample design when building a regression tree using a recursive partitioning algorithm. Sufficient conditions are established for asymptotic design L 2 consistency of these regression trees as estimators for an arbitrary regression function. The proposed method is illustrated with Occupational Employment Statistics establishment survey data linked to Quarterly Census of Employment and Wage payroll data of the Bureau of Labor Statistics. Performance of the nonparametric estimator is investigated through a simulation study based on this example.

stat.ME↗

arXiv2012

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

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.

stat.AP↗