VARIANCE ESTIMATION IN TWO-PHASE SAMPLING

Two‐phase sampling is often used for estimating a population total or mean when the cost per unit of collecting auxiliary variables, x, is much smaller than the cost per unit of measuring a characteristic of interest, y. In the first phase, a large sample s₁ is drawn according to a specific sampling design p(s₁) , and auxiliary data x are observed for the units i∈s₁ . Given the first‐phase sample s₁ , a second‐phase sample s₂ is selected from s₁ according to a specified sampling design {p(s₂∣s₁) } , and (y, x) is observed for the units i∈s₂ . In some cases, the population totals of some components of x may also be known. Two‐phase sampling is used for stratification at the second phase or both phases and for regression estimation. Horvitz–Thompson‐type variance estimators are used for variance estimation. However, the Horvitz–Thompson ( Horvitz & Thompson, J. Amer. Statist. Assoc. 1952 ) variance estimator in uni‐phase sampling is known to be highly unstable and may take negative values when the units are selected with unequal probabilities. On the other hand, the Sen–Yates–Grundy variance estimator is relatively stable and non‐negative for several unequal probability sampling designs with fixed sample sizes. In this paper, we extend the Sen–Yates–Grundy ( Sen , J. Ind. Soc. Agric. Statist. 1953; Yates & Grundy , J. Roy. Statist. Soc. Ser. B 1953) variance estimator to two‐phase sampling, assuming fixed first‐phase sample size and fixed second‐phase sample size given the first‐phase sample. We apply the new variance estimators to two‐phase sampling designs with stratification at the second phase or both phases. We also develop Sen–Yates–Grundy‐type variance estimators of the two‐phase regression estimators that make use of the first‐phase auxiliary data and known population totals of some of the auxiliary variables.     

Stochastic comparisons and properties of conditional generalized order statistics

Generalized order statistics introduced by Kamps [1995. A concept of generalized order statistics. J. Statist. Plann. Inference 48, 1–23] provides a unified approach to a variety of models concerning ordered random variables. This paper carries out stochastic comparisons of conditional generalized order statistics in terms of the likelihood ratio order and the hazard rate order from two samples, and establishes some stochastic monotonicity properties. The main results strengthen and generalize the corresponding results established recently in the literature. Finally, some applications of the main results are given as well.     

Empirical likelihood confidence intervals under imputation for missing survey data from stratified simple random sampling

Missing observations due to non‐response are commonly encountered in data collected from sample surveys. The focus of this article is on item non‐response which is often handled by filling in (or imputing) missing values using the observed responses (donors). Random imputation (single or fractional) is used within homogeneous imputation classes that are formed on the basis of categorical auxiliary variables observed on all the sampled units. A uniform response rate within classes is assumed, but that rate is allowed to vary across classes. We construct confidence intervals (CIs) for a population parameter that is defined as the solution to a smooth estimating equation with data collected using stratified simple random sampling. The imputation classes are assumed to be formed across strata. Fractional imputation with a fixed number of random draws is used to obtain an imputed estimating function. An empirical likelihood inference method under the fractional imputation is proposed and its asymptotic properties are derived. Two asymptotically correct bootstrap methods are developed for constructing the desired CIs. In a simulation study, the proposed bootstrap methods are shown to outperform traditional bootstrap methods and some non‐bootstrap competitors under various simulation settings. The Canadian Journal of Statistics 47: 281–301; 2019 © 2019 Statistical Society of Canada     

Daily perceptions of relationship quality with leaders: implications for follower well-being

ABSTRACT

We examined how a key relationship at work, an employee’s relationship with their leader, affects employee daily well-being. In a study of 129 employees across a variety of industries, we examined how follower perceptions of their daily leader–member exchange (LMX) quality across a workweek influenced their well-being (n?=?468 observations). Results provided general support for our hypotheses. Specifically, we found that on days when employees perceived a higher quality LMX relationship with their leader, they were more likely to report a sense of belongingness, which was then positively associated with daily reports of vigor and negatively associated with emotional exhaustion. Lagged analyses showed that perceptions of LMX quality were also negatively associated with reports of emotional exhaustion the following workday suggesting that these effects may persist over time. Finally, we found that day-to-day variation in reports of LMX quality attenuated the beneficial effects of LMX on relatedness and vigor supporting our hypothesis that uncertainty related to resource availability may contribute to a threat mindset focused on resource conservation rather than engagement. Implications and future research on leadership and employee well-being are discussed.     

Diaspora and mapping methodologies: tracing transnational digital connections with ‘mattering maps’

There is a methodological tendency in work on diaspora and digital media for quantitative investigations to approach diaspora in static ways that contrast with theories of diaspora as a dynamic cultural formation. On the other hand, qualitative, ethnographic work tends not to engage with digital methods and quantitative data‐driven investigation. In this article, we sketch this methodological and disciplinary disconnect and address it by proposing a model for understanding digitally mediated formations of diaspora that combines digital methods techniques with a sensitivity to ethical and theoretical discussions of migration and diaspora. Drawing on interpretive epistemologies and feminist research ethics, we present a case study analysis of a locally informed, Turkish–Dutch issue. We argue for a method that produces ‘mattering maps’. This involves tracking and visualizing digital traces of an issue across web platforms (Google Search results, Facebook pages, and Instagram posts) and integrating this with an analysis of the face‐to‐face interview responses of a key issue actor.     

A Screening,Brief Intervention,and Referral to Treatment Collaborative: Impact for Social Work Students

Screening brief intervention, and referral to treatment (SBIRT) is a clinical intervention empirically linked to reductions in substance usage. However, less is known on how to effectively train social work students on SBIRT. This study examines the impact of a SBIRT training administered to BSW and MSW students (N=228) at the University of Oklahoma. Based on exposure to the training, our evaluation measured changes on students’ objective knowledge of how to perform SBIRT and in attitudes toward practicing with substance users. Results indicated exposure to the training was associated with robust gains across variables and across classes. Such results suggest that the proffered SBIRT training is an effective means to educate social work students on substance abuse practice.     

Quite an encumbrance? Work-related obstacles to Finnish fathers’ take-up of parental leave

ABSTRACT

An ideal of involved fatherhood has become popular in the developed countries, but even countries like Finland that have introduced an individual father’s quota to parental leave are far from gender equality in parents’ leave practices. Lack of support and negative attitudes at workplaces or pressure at work are among the obstacles to fathers’ leave take-up. The study used survey data from fathers and interviews at workplaces to explore the role of work and workplaces among the many aspects related to fathers’ possibilities of taking leave. The results showed that fathers’ income and workplace characteristics were associated with taking leave. Few fathers mentioned employers’ objection as a hinder. Father’s recent unemployment and anticipated difficulty of taking a long time off played a more important role. At workplaces, the obstacles to taking long leave were related to fathers’ ideals about a committed worker and to the nature and organizing of work. Additionally, the leave exceeding fathers’ quota might not be understood as ‘for fathers’. The spouse’s situation and fathers’ gendered perceptions about parental responsibilities were also important for leave practices. The findings suggest that policy development towards a longer father’s quota could make fathers’ care responsibilities visible also at workplaces.     

Estimation of a rare sensitive attribute for two-stage randomized response model in probability proportional to size sampling using Poisson probability distribution

This paper addresses the estimation of the mean number of individuals in the population who possess a rare sensitive attribute using Poisson distribution for the situations of (i) clustered population and (ii) stratified population with clusters are strata units. Properties of the proposed estimation procedures have been discussed when the proportion of a rare unrelated non-sensitive attribute is assumed to be known as well as unknown. Empirical studies are carried out to support the theoretical results which showed dominance over Lee et al. [Estimation of a rare sensitive attribute in probability proportional to size measures using Poisson distribution. Statistics (Ber). 2014;48(3):685–709] estimation procedures.     

“The Panhandle is Different than the Peninsula”: How Rural Colleges in Florida Implemented Education Reform

When it comes to the creation of higher education policy, state legislators are challenged with addressing the diverse academic needs of college students enrolled across a spectrum of institutions. In this paper, we explore how 203 individuals at five rural-serving colleges in Florida engaged in state-wide developmental education (DE) reform using the theoretical framework of “situated cognition.” Specifically, we ask whether the colleges implemented DE reform in unique ways that may have differed from their non-rural counterparts and what their rationale was for doing so. Our work indicates that institutional culture and capacity generated a combination of strengths and constraints in how reform played out in a rural context. In some ways, a rural-serving identity made colleges more flexible and better able to adjust their advising and curricular structures to comply with the mandates of SB 1720. But in other ways, rural-serving institutions lacked the resources necessary to adequately support students and campus personnel through this dramatic transition toward new ways of providing DE. In looking toward the passage and implementation of future education reform, we call upon policymakers to consider more deeply the design of policies to ensure they support rural and non-rural colleges alike.