Judgment Post-stratification Research Articles

Analysis of Ordinal Populations from Judgment Post-Stratification

In surveys requiring cost efficiency, such as medical research, measuring the variable of interest (e.g., disease status) is expensive and/or time-consuming; however, we often have access to easily obtainable characteristics about sampling units. These characteristics are not typically employed in the data collection process. Judgment post-stratification (JPS) sampling enables us to supplement the random samples from the population of interest with these characteristics as ranking information. This paper develops methods based on the JPS samples for estimating categorical ordinal populations. We develop various estimators from the JPS data even for situations where the JPS suffers from empty strata. We also propose the JPS estimators using multiple ranking resources. Through extensive numerical studies, we evaluate the performance of the methods in estimating the population. Finally, the developed estimation methods are applied to bone mineral data to estimate the bone disorder status of women aged 50 and older.

Another look at regression analysis using ranked set samples with application to an osteoporosis study.

Statistical learning with ranked set samples has shown promising results in estimating various population parameters. Despite the vast literature on rank-based statistical learning methodologies, very little effort has been devoted to studying regression analysis with such samples. A pressing issue is how to incorporate the rank information of ranked set samples into the analysis. We propose two methodologies based on a weighted least squares approach and multilevel modeling to better incorporate the rank information of such samples into the estimation and prediction processes of regression-type models. Our approaches reveal significant improvements in both estimation and prediction problems over already existing methods in the literature and the corresponding ones with simple random samples. We study the robustness of our methods with respect to the misspecification of the distribution of the error terms. Also, we show that rank-based regression models can effectively predict simple random test data by assigning ranks to them a posteriori using judgment poststratification. Theoretical results are augmented with simulations and an osteoporosis study based on a real data set from the Bone Mineral Density (BMD) program of Manitoba to estimate the BMD level of patients using easy to obtaincovariates.

Efficient estimation of the odds using judgment post stratification

This work deals with problem of estimating the odds using judgment post stratification (JPS) sampling design. Several estimators of the odds are described and the asymptotic normality of each of them is established. Monte Carlo simulation study is then used to compare different estimators of the odds in the JPS with the standard estimator in simple random sampling (SRS) with replacement for both perfect/imperfect ranking and for both JPS data with/without empty strata. The comparison results indicate that the estimators developed here can be highly more efficient than their SRS counterpart in some certain circumstances. Finally, a real dataset from the third National Health and Nutrition Examination Survey (NHANES III) is employed for illustration purposes.

Bootstrap Methods for Judgment Post Stratification

It has been shown in the literature that judgment post stratification (JPS) sampling design often leads to more efficient statistical inference than what is possible to obtain in simple random sampling (SRS) design of comparable size. Since the JPS is a cost-efficient sampling design, a large enough sample size may not be available to use normal theory of the estimators. In this paper, we describe two bootstrap methods for JPS sampling scheme, one of which has been already used in the literature without studying its consistency and the other is new. We also show that both bootstrap approaches are consistent. We then investigate the use of the bootstrap methods for constructing confidence intervals for the population mean and compare them with the confidence interval of the population mean obtained via normal approximation (NA) method using Monte Carlo simulation. It is found that for the asymmetric distributions, one of the bootstrap methods we describe in the paper often leads to a closer coverage probability (CP) to the nominal level than NA method. Finally, a real dataset is analysed for illustration.

Judgment post-stratification in finite mixture modeling: An example in estimating the prevalence of osteoporosis.

Judgment post-stratification is used to supplement observations taken from finite mixture models with additional easy to obtain rank information and incorporate it in the estimation of model parameters. To do this, sampled units are post-stratified on ranks by randomly selecting comparison sets for each unit from the underlying population and assigning ranks to them using available auxiliary information or judgment ranking. This results in a set of independent order statistics from the underlying model, where the number of units in each rank class is random. We consider cases where one or more rankers with different ranking abilities are used to provide judgment ranks. The judgment ranks are then combined to produce a strength of agreement measure for each observation. This strength measure is implemented in the maximum likelihood estimation of model parameters via a suitable expectation maximization algorithm. Simulation studies are conducted to evaluate the performance of the estimators with or without the extra rank information. Results are applied to bone mineral density data from the third National Health and Nutrition Examination Survey to estimate the prevalence of osteoporosis in adult women aged 50 and over.

An omnibus two-sample test for ranked-set sampling data

We develop an omnibus two-sample test for ranked-set sampling (RSS) data. The test statistic is the conditional probability of seeing the observed sequence of ranks in the combined sample, given the observed sequences within the separate samples. We compare the test to existing tests under perfect rankings, finding that it can outperform existing tests in terms of power, particularly when the set size is large. The test does not maintain its level under imperfect rankings. However, one can create a permutation version of the test that is comparable in power to the basic test under perfect rankings and also maintains its level under imperfect rankings. Both tests extend naturally to judgment post-stratification, unbalanced RSS, and even RSS with multiple set sizes. Interestingly, the tests have no simple random sampling analog.

Inference on a distribution function from ranked set samples

Consider independent observations $$(X_i,R_i)$$ with random or fixed ranks $$R_i$$, while conditional on $$R_i$$, the random variable $$X_i$$ has the same distribution as the $$R_i$$-th order statistic within a random sample of size k from an unknown distribution function F. Such observation schemes are well known from ranked set sampling and judgment post-stratification. Within a general, not necessarily balanced setting we derive and compare the asymptotic distributions of three different estimators of the distribution function F: a stratified estimator, a nonparametric maximum-likelihood estimator and a moment-based estimator. Our functional central limit theorems generalize and refine previous asymptotic analyses. In addition, we discuss briefly pointwise and simultaneous confidence intervals for the distribution function with guaranteed coverage probability for finite sample sizes. The methods are illustrated with a real data example, and the potential impact of imperfect rankings is investigated in a small simulation experiment.

Estimation of population proportion for judgment post-stratification

This paper is concerned with the problem of estimating a population proportion p in a judgment post-stratification (JPS) sampling scheme. Different proportion estimators are considered, among which some are specifically designed to deal with JPS samples with empty strata; and asymptotic normality is established for each. A Monte Carlo simulation study and two examples using data from medical studies are employed to examine the performance of these proportion estimators under both perfect and imperfect ranking and for JPS data both with and without empty strata. It is shown that the JPS scheme improves estimation of the population proportion in a very wide range of settings as compared to simple random sampling (SRS). Also, findings about the relative performance of the different estimators are provided to help practitioners determine which estimator should be used under certain situations.

Entropy estimation from judgment post stratified data

This article concerns entropy estimation using judgment post stratification sampling. Some nonparametric estimators are developed and shown to be consistent. Monte Carlo simulations are used to compare these estimators with their competitors in simple random sampling. The results indicate the preference of the new estimators.

Unified ranked sampling

A class of ranked sample methodologies is presented, which unifies the existing methods of ranked set sampling and judgment post-stratification. This general class contains many new, promising methods and allows future ranked sampling techniques to be developed with greater scope.

Some nonparametric tests of perfect judgment ranking for judgment post stratification

We develop some nonparametric tests of perfect judgment ranking for judgment post stratification sampling scheme. We show that the best proposed test beats the best existing nonparametric test of perfect judgment ranking in ranked set sampling applied to the judgment post stratification case by conditioning on the observed stratum sizes.

Some theoretical results concerning non parametric estimation by using a judgment poststratification sample

ABSTRACTIn this paper, some of the properties of non parametric estimation of the expectation of g(X) (any function of X), by using a judgment poststratification sample (JPS), have been discussed. A class of estimators (including the standard JPS estimator and a JPS estimator proposed by Frey and Feeman (2012, Comput. Stat. Data An.) is considered. The paper provides mean and variance of the members of this class, and examines their consistency and asymptotic distribution. Specifically, the results are for the estimation of population mean, population variance, and cumulative distribution function. We show that any estimators of the class may be less efficient than simple random sampling (SRS) estimator for small sample sizes. We prove that the relative efficiency of some estimators in the class with respect to balanced ranked set sampling (BRSS) estimator tends to 1 as the sample size goes to infinity. Furthermore, the standard JPS mean estimator, and Frey–Feeman JPS mean estimator are specifically studied and we show that two estimators have the same asymptotic distribution. For the standard JPS mean estimator, in perfect ranking situations, optimum values of H (the ranking class size), for different sample sizes, are determined non parametrically for populations that are not heavily skewed or thick tailed.

An isotonized variance estimator for judgment post stratified data

In this paper, we develop a nonparametric variance estimator for judgment post stratification sampling scheme. The estimator is obtained by using the fact that the distributions of post strata are often stochastically ordered. We prove that our estimator is consistent. Our simulation results indicate that the proposed estimator has good performance in comparison with its leading competitors in the literature.

Variance estimation in ranked set sampling using a concomitant variable

We propose a nonparametric variance estimator when ranked set sampling (RSS) and judgment post stratification (JPS) are applied by measuring a concomitant variable. Our proposed estimator is obtained by conditioning on observed concomitant values and using nonparametric kernel regression.

Generalized Isotonized Mean Estimators for Judgment Post-stratification with Multiple Rankers

In this paper, we propose a set of new mean estimators for judgment post-stratified data with multiple rankers. The new estimators take into account matrix partial ordering in cumulative distribution functions of rank strata, and they are derived by improving existing estimators through employing the order constraints and solving a generalized isotonic regression problem. Numerical studies show that the proposed isotonized mean estimators outperform the existing estimators. Finally, the proposed estimators are applied to estimating the average tree height using the tree data in Chen et al. (Ranked set sampling: theory and applications, Springer, New York, 2006).

Variance estimation using judgment post-stratification

We consider the problem of estimating the variance of a population using judgment post-stratification. By conditioning on the observed vector of ordered in-stratum sample sizes, we develop a conditionally unbiased nonparametric estimator that outperforms the sample variance except when the rankings are very poor. This estimator also outperforms the standard unbiased nonparametric variance estimator from unbalanced ranked-set sampling.

An improved mean estimator for judgment post-stratification

We prove that the standard nonparametric mean estimator for judgment post-stratification is inadmissible under squared error loss within a certain class of linear estimators. We derive alternate estimators that are admissible in this class, and we show that one of them is always better than the standard estimator. The reduction in mean squared error from using this alternate estimator can be as large as 10% for small set sizes and small sample sizes.

Isotonized CDF Estimation from Judgment Poststratification Data with Empty Strata

In applications that require cost efficiency, sample sizes are typically small so that the problem of empty strata may often occur in judgment poststratification (JPS), an important variant of balanced ranked set sampling. In this article, we consider estimation of population cumulative distribution functions (CDF) from JPS samples with empty strata. In the literature, the standard and restricted CDF estimators (Stokes and Sager, 1988, Journal of the American Statistical Association 83, 374381; Frey and Ozturk, 2011, Annals of the Institute of Statistical Mathematics, to appear) do not perform well when simply ignoring empty strata. In this article, we show that the original isotonized estimator (Ozturk, 2007, Journal of Nonparametric Statistics 19, 131-144) can handle empty strata automatically through two methods, MinMax and MaxMin. However, blindly using them can result in undesirable results in either tail of the CDF. We thoroughly examine MinMax and MaxMin and find interesting results about their behaviors and performance in the presence of empty strata. Motivated by these results, we propose modified isotonized estimators to improve estimation efficiency. Through simulation and empirical studies, we show that our estimators work well in different regions of the CDF, and also improve the overall performance of estimating the whole function.

Constrained nonparametric estimation of the mean and the CDF using ranked-set sampling with a covariate

Ranked-set sampling (RSS) and judgment post-stratification (JPS) are related schemes in which more efficient statistical inference is obtained by creating a stratification based on ranking information. The rankings may be completely subjective, or they may be based on values of a covariate. Recent work has shown that regardless of how the rankings are done, the in-stratum cumulative distribution functions (CDFs) must satisfy certain constraints, and we show here that if the rankings are done according to a covariate, then tighter constraints must hold. We also show that under a mild stochastic ordering assumption, still tighter constraints must hold. Taking advantage of these new constraints leads to improved small-sample estimates of the in-stratum CDFs in all RSS and JPS settings. For JPS, the new constraints also lead to improved estimates of the overall CDF and the population mean.

A note on ranked-set sampling using a covariate

Ranked-set sampling (RSS) and judgment post-stratification (JPS) use ranking information to obtain more efficient inference than is possible using simple random sampling. Both methods were developed with subjective, judgment-based rankings in mind, but the idea of ranking using a covariate has received a lot of attention. We provide evidence here that when rankings are done using a covariate, the standard RSS and JPS mean estimators no longer make efficient use of the available information. We first show that when rankings are done using a covariate, the standard nonparametric mean estimators in JPS and unbalanced RSS are inadmissible under squared error loss. We then show that when rankings are done using a covariate, nonparametric regression techniques yield mean estimators that tend to be significantly more efficient than the standard RSS and JPS mean estimators. We conclude that the standard estimators are best reserved for settings where only subjective, judgment-based rankings are available.