Search

Abstract--Pooling consumption data from different countries for demand system estimation is attractive because it increases both sample size and the range of variation of relative prices and income. The major objection to pooling is that different countries may have different demand system parameters. This paper proposes and estimates specifications that permit pooling while allowing both short-run and long-run demand systems to differ across countries. Using data from Belgium, the United Kingdom and the United States, we find that, although pooling was accepted for some pairs of countries and some specifications, it was rejected for most. We conclude that caution is appropriate in pooling international consumption data. [ABSTRACT FROM AUTHOR]

The conditions under which interdisciplinary research helps improve research performance are explored in a heterogeneous sample of 67 ongoing academic interdisciplinary teams from the United States and Canada. We conclude that interdisciplinary research is more appropriate for very difficult research questions and at early stages of the research process. Interdisciplinary research was found to inhibit the outputs of articles, books, papers, and technical reports when the research had a clearly identified client prior to the start of the project. Interdisciplinary research was not found to be more helpful for either applied or pure research. It was not found to be helpful either at the concluding stage of research or when used throughout the research process. [ABSTRACT FROM AUTHOR]

A proper analysis of data from complex sample surveys requires special consideration for estimating standard errors. Special techniques and software packages are available, including Taylor series linearization (delta method), balanced repeated replication, and jackknife. Before their use, it is often necessary to make certain modifications in original data structure, to conform to computing method requirements. The most common modification is to form pseudostrata by collapsing substrata or partitioning a string of geographic clusters. This paper examines the performance of the delta method when it is applied to a complex community survey data set in which sequentially drawn clusters of households are partitioned to form pseudostrata. Standard errors of rates, regression coefficients, and odds ratios are compared with those computed from the variation of replicates built into the sample design. The results demonstrate that an analysis of complex survey data should use an appropriate method for estimating standard errors, and that pseudostrata would produce reasonable estimates of standard errors for rates and regression coefficients, with mixed results for odds ratios. [ABSTRACT FROM AUTHOR]

The relationship between the number of road accidents and the volume of traffic is econometrically estimated in this paper. A stratified random sample of 399 road segments covering urban and rural roads in New York State (USA) in 1985 is analyzed. No significant externality is detected. On very high volume urban roads the marginal to average accident rate ratio is 1.06. On all other roads, a ratio of 0.98 is estimated. These findings contrast sharply with those of Vickrey, who estimated a ratio of 1.5 for California freeways in 1962. The results are, however, consistent with the 'official view' of both the US and UK highway authorities that no significant accident externality exists. [ABSTRACT FROM AUTHOR]

This paper uses the econometrics of sample selection to estimate the net wage differential between workers in the public and private sectors of the United States. The results indicate workers self select by systematically locating in the sector which offers them the higher expected wage. As a consequence, previous research based on the assumption of random allocation between sectors is subject to serious specification bias. The present estimation indicates government workers are slightly underpaid taking into account their observed characteristics and the selection process. [ABSTRACT FROM AUTHOR]

This paper develops methods for constructing asymptotically valid confidence intervals for the date of a single break in multivariate time series, including I(0), I(1), and deterministically trending regressors. Although the width of the asymptotic confidence interval does not decrease as the sample size increases, it is inversely related to the number of series which have a common break date, so there are substantial gains to multivariate inference about break dates. These methods are applied to two empirical examples: the mean growth rate of output in three European countries, and the mean growth rate of U.S. consumption, investment, and output. [ABSTRACT FROM AUTHOR]

In applied tax and income distribution studies, the predominant practice Is to use the Lorenz curve and one or morn indexes derived from the Lorenz curve to describe the distributional effects of tax law changes. However where sample data are used, sampling variation is present. Observed distributional differences across alternative tax systems may, therefore, be attributable to sampling variability. This paper outlines and explains the methodology by which data underlying Lorenz curves can be used for Inferential rather than Just descriptive purposes. The methodology is demonstrated by applying it to evaluate the effect of the income tax system on income inequality and that of the earned income credit on overall progressivity and income inequality. [ABSTRACT FROM AUTHOR]

This article reports on the study conducted in the United States to find out whether effective newspaper sampling is possible or not. To select the sample purposively in two equal increments of 50 newspapers each, the universe was stratified with respect to certain objective characteristics, the characteristics of an ideal 50 paper sub-sample were determined, and two sub-samples were selected deliberately with characteristics matching those of the ideal sub-sample. These were then coded for certain items of military interest, and results from the two compared. In conducting the study, the sampling universe was limited to general circulation. Then the two samples were compared with the help of correlation method. Correlation between the content of the two samples was close enough to justify the conclusion that effective newspaper sampling is possible. It is based upon the consistent, day-to-day, subject-by-subject correlation between the two sub-samples.

In recent years, there has been a proliferation of voluntary citizens' patrols in the urban areas. This paper focuses on a citizens' patrol of the relatively enduring type: the Indian Patrol of Minneapolis, Minnesota, which was sponsored by the American Indian Movement (AIM). The material presented here is drawn from a field study which was conducted between January 1969 and June 1970. Patrols were observed by the author on twenty evenings. In addition, the research included a survey of a selected sample of 40 AIM members and a survey of a random sample of 44 residents of the neighborhood served by the Indian Patrol. The results of the neighborhood survey support the notion that the Indian Patrol in the East Franklin Avenue neighborhood functioned primarily as a means by which a particular ideology could be demonstrated by one group and perceived by another. Although almost half of the residents interviewed had heard about the Indian Patrol, very few had ever seen it; even fewer had any personal experience with it. None seemed to know how to get in touch with the patrol in time of need.

Hill's typology of comparative family research is a challenging attempt to classify international, e.g., cross-cultural, research. This paper intends to criticize some details of this classification, particularly concerning methodological issues. The main point stressed here is the problem of the reconceptualization of the research problem for each different country in which the research has to be carried out. [ABSTRACT FROM AUTHOR]

This paper studies the design effect of counting rules, for linking deaths to housing units where they are enumerated in the survey, on the sampling variance of dual system and single system estimators of death registration completeness. It investigates estimators based on conventional rules that uniquely link each death to a single housing unit as well as estimators based on multiplicity rules which permit deaths to be linked to more than one housing unit. Sampling variance formulas are derived containing parameters that reflect the efficiency of the counting rule. Estimates of these parameters for different counting rules are compared utilizing information that was collected in a mortality survey experiment. Finally, the design of a national death registration test is considered and the sample size implications of different counting rules are compared. [ABSTRACT FROM AUTHOR]

Home health care is a rapidly growing healthcare sector worldwide. Home health professionals face unique challenges related to preventing and controlling infections, which are likely to amplify during an infectious disease outbreak (e.g. SARS-CoV-2). Little is known about the current state of infection prevention and control-related policies and outbreak preparedness at U.S. home health agencies. In this study, we conducted a national survey to assess infection prevention and control-related policies, infrastructure, and procedures prior to the SARS-CoV-2 pandemic. Cross-sectional study. Using a stratified random sample of 1506 U.S. home health agencies, we conducted a 61-item survey (paper and online) from November 9, 2018 to December 31, 2019. Survey data were linked to publicly-available data on the quality of patient care, patient satisfaction, and other agency characteristics. Probability weights were developed to account for sample design and nonresponse; Pearson's χ2, Fisher's exact, t-tests or linear regression were used to compare the universe of agencies/respondents and urban/rural agencies. 35.6% of agencies responded (n = 536). Most home health personnel in charge of infection prevention and control have other responsibilities; one-third have no formal infection prevention and control training. Rural agencies are more likely to not have anyone in charge of infection prevention and control compared to those in urban areas. About 22% of agencies implement recommended guidelines when administering antibiotics. Less than a third (26.4%) report that their staff vaccination rates were higher than 95% during the last flu season. Only 48.1% of agencies accept patients requiring ventilation, and of those, 40.9% located in rural areas do not have specific infection prevention and control policies for ventilated patients, compared to 20.8% in urban areas (p < 0.001). Only 39.7% of agencies provide N95 respirators to their clinical staff; rural agencies are significantly more likely to provide those supplies than urban agencies (50.7% vs. 37.7%, p = 0.004). Lastly, agencies report their greatest challenges with infection prevention and control are collecting/reporting infection data and adherence to/monitoring of nursing bag technique. Prior to the SARS-CoV-2 pandemic, we found that infection prevention and control was suboptimal among U.S. home health care agencies. Consequently, most agencies have limited capacity to respond to infectious disease outbreaks. Staff and personal protective equipment shortages remain major concerns, and agencies will need to quickly adjust their existing infection prevention and control policies and potentially create new ones. In the long-term, agencies also need to improve influenza vaccination coverage among their staff. Tweetable abstract: Infection prevention and control infrastructure, policies and procedures and outbreak preparedness at U.S. home health agencies was found to be suboptimal in nationally-representative survey conducted just prior to the COVID-19 pandemic. [ABSTRACT FROM AUTHOR]

Many low and middle-income countries in Asia experience greater food availability as their economies grow, potentially solving previous problems of undernutrition. However, economic growth may not sufficiently reduce undernutrition and instead increase over nutrition, creating a double burden of malnutrition. In this paper we investigate one of the most prevalent and seemingly paradoxical manifestations of the double burden, a stunted child and an overweight mother (SCOM) living in the same household. We use household-level data from 11 Asian countries from 2005 to 2018, as well as macroeconomic data on the Gross Domestic Product (GDP) and the GDP growth of these countries to estimate which household-level and macroeconomic factors predict SCOM. We also determine the relative predictive power of our variables for SCOM and investigate if predictors of SCOM also predict other forms of malnutrition. We find that the predictive power of household-level factors for SCOM is comparatively more important than that of macroeconomic variables. Additionally, we explore the effect of including population subgroup-level trends in maternal overweight and child stunting on the predictive power of our macro-level and household-level variables. The increasing prevalence of SCOM is driven by a rise in maternal overweight, coupled with stagnant rates of child stunting, but remains a unique phenomenon that differs from other indicators of malnutrition. This supports explanations that suggest increasing consumption of high-energy but low-nutrient foods is the underlying cause of SCOM, leading to weight gain in adults and stunting in under-five children. • Increasing rates of Asian households with overweight mothers and stunted children. • A nutritional transition in the wake of economic growth an often-conjectured driver. • Known household-level determinants confirmed in large cross-country dataset. • Both growth and household-level determinants are predictors of the double burden. • Household-level factors with greater direct predictive power than economic growth. [ABSTRACT FROM AUTHOR]

Recalls statistical research papers previously published in the United States. Survey research; Sampling; Technical problems of statisticians at the U.S. Census Bureau.

The article comments on the paper made by the American Psychological Association Task Force on Statistical Inference, which emphasized the importance of statistical power for the design and reporting of research, in the U.S. According to the author, the Task Force on Statistical Inference may have developed an unintended false impression that increasing sample size is the best solution for improving statistical power. The author suggests that psychologists must take into consideration the design methods rather than augmenting sample size to reduce cost and lessen the burden on the subjects.

The article presents an abstract of an article on the statistical methodology of the National Jewish Population Survey in the U.S. This paper reports on the three aspects of the recently conducted study. The first section outlines the survey's sample design detailing the complexities of sampling American Jewry. The second section reports the response rates and sample weights. While the third section estimated the American Jewish population.