Kristen Olson

The university of Michigan dioxin exposure study: Predictors of human serum dioxin concentrations in Midland and Saginaw, Michigan

Background We conducted a population-based human exposure study in response to concerns among the population of Midland and Saginaw counties, Michigan, that discharges by the Dow Chemical Company of dioxin-like compounds into the nearby river and air had led to an increase in residents’ body burdens of polychlorinated dibenzofurans (PCDDs), polychlorinated dibenzofurans (PCDFs), and dioxin-like polychlorinated biphenyls (PCBs), here collectively referred to as “dioxins.” Objectives We sought to identify factors that explained variation in serum dioxin concentrations among the residents of Midland and Saginaw counties. Exposures to dioxins in soil, river sediments, household dust, historic emissions, and contaminated fish and game were of primary interest. Methods We studied 946 people in four populations in the contaminated area and in a referent population, by interview and by collection of serum, household dust, and residential soil. Linear regression was used to identify factors associated with serum dioxins. Results Demographic factors explained a large proportion of variation in serum dioxin concentrations. Historic exposures before 1980, including living in the Midland/Saginaw area, hunting and fishing in the contaminated areas, and working at Dow, contributed to serum dioxin levels. Exposures since 1980 in Midland and Saginaw counties contributed little to serum dioxins. Conclusions This study provides valuable insights into the relationships between serum dioxins and environmental factors, age, sex, body mass index, smoking, and breast-feeding. These factors together explain a substantial proportion of the variation in serum dioxin concentrations in the general population. Historic exposures to environmental contamination appeared to be of greater importance than recent exposures for dioxins.

Multiple auxiliary variables in nonresponse adjustment

Prior work has shown that effective survey nonresponse adjustment variables should be highly correlated with both the propensity to respond to a survey and the survey variables of interest. In practice, propensity models are often used for nonresponse adjustment with multiple auxiliary variables as predictors. These auxiliary variables may be positively or negatively associated with survey participation, they may be correlated with each other, and can have positive or negative relationships with the survey variables. Yet the consequences for nonresponse adjustment of these conditions are not known to survey practitioners. Simulations are used here to examine the effects of multiple auxiliary variables with opposite relationships with survey participation and the survey variables. The results show that bias and mean square error of adjusted respondent means are substantially different when the predictors have relationships of the same directions compared to when they have opposite directions with either propensity or the survey variables. Implications for nonresponse adjustment and responsive designs will be discussed.

An Examination of Questionnaire Evaluation by Expert Reviewers

Expert reviews are frequently used as a questionnaire evaluation method but have received little empirical attention. Questions from two surveys are evaluated by six expert reviewers using a standardized evaluation form. Each of the questions has validation data available from records. Large inconsistencies in ratings across the six experts are found. Despite the lack of reliability, the average expert ratings successfully identify questions that had higher item nonresponse rates and higher levels of inaccurate reporting. This article provides empirical evidence that experts are able to discern questions that manifest data quality problems, even if individual experts vary in what they rate as being problematic. Compared to a publicly available computerized question evaluation tool, ratings by the human experts positively predict questions with data quality problems, whereas the computerized tool varies in success in identifying these questions. These results indicate that expert reviews have value in identifying question problems that result in lower survey data quality.

The University of Michigan Dioxin Exposure Study: Population Survey Results and Serum Concentrations for Polychlorinated Dioxins, Furans, and Biphenyls

Background The University of Michigan Dioxin Exposure Study was undertaken to address concerns that the discharge of polychlorinated dibenzo-p-dioxins (PCDDs) and polychlorinated dibenzo furans (PCDFs) from the Dow Chemical Company in the Midland, Michigan, area had resulted in contamination of soils in the Tittabawassee River floodplain and the city of Midland, leading to an increase in residents’ body burdens of these compounds. Objective In this article we present descriptive statistics from the resident survey and sampling of human serum, household dust, and soil and compare them with other published values. Methods From a multistage random sample of populations in four areas of Midland and Saginaw counties and from a distant referent population, we interviewed 946 adults, who also donated blood for analysis of PCDDs, PCDFs, and polychlorinated biphenyls (PCBs). Samples of household dust and house perimeter soil were collected from consenting subjects who owned their property. Results All five study populations were comparable in age, race, sex, and length of residence in their current home. Regional differences existed in employment history, personal contact with contaminated soils, and consumption of fish and game from contaminated areas. Median soil concentrations were significantly increased around homes in the Tittabawassee River floodplain (11.4 ppt) and within the city of Midland (58.2 ppt) compared with the referent population (3.6 ppt). Median serum toxic equivalencies were significantly increased in people who lived in the floodplain (23.2 ppt) compared with the referent population (18.5 ppt). Conclusions Differences in serum dioxin concentrations among the populations were small but statistically significant. Regression modeling is needed to identify whether the serum concentrations of PCDDs, PCDFs, and PCBs are associated with contaminated soils, household dust, and other factors.

The University of Michigan Dioxin Exposure Study: Methods for an Environmental Exposure Study of Polychlorinated Dioxins, Furans, and Biphenyls

Background The University of Michigan Dioxin Exposure Study (UMDES) was undertaken in response to concerns that the discharge of dioxin-like compounds from the Dow Chemical Company facilities in Midland, Michigan, resulted in contamination of soils in the Tittabawassee River floodplain and areas of the city of Midland, leading to an increase in residents’ body burdens of polychlorinated dibenzodioxins and polychlorinated dibenzofurans. Objectives The UMDES is a hypothesis-driven study designed to answer important questions about human exposure to dioxins in the environment of Midland, where the Dow Chemical Company has operated for > 100 years, and in neighboring Saginaw, Michigan. In addition, the UMDES includes a referent population from an area of Michigan in which there are no unusual sources of dioxin exposure and from which inferences regarding the general Michigan population can be derived. A central goal of the study is to determine which factors explain variation in serum dioxin levels and to quantify how much variation each factor explains. Conclusions In this article we describe the study design and methods for a large population-based study of dioxin contamination and its relationship to blood dioxin levels. The study collected questionnaire, blood, dust, and soil samples on 731 people. This study provides a foundation for understanding the exposure pathways by which dioxins in soils, sediments, fish and game, and homegrown produce lead to increased body burdens of these compounds.

Paradata for Nonresponse Adjustment

Survey researchers and practitioners use nonresponse adjustment weights to mitigate the effects of survey nonresponse on sample estimates. One challenge in creating these weights is finding useful auxiliary data that predict both the probability of participating in the survey and the survey variables of interest. This article reviews the use of paradata for nonresponse adjustment. Five different types of paradata are considered: neighborhood observations, observations of the sampled housing unit, observations of persons in the sampled housing unit, call records, and observations about the interviewer-householder interaction. Empirical evidence about the predictive value of these paradata for predicting both participation and survey variables is examined. Challenges of using these paradata are also identified, along with outstanding issues and opportunities related to the use of paradata for nonresponse adjustment.

Accuracy of Within-household Selection in Web and Mail Surveys of the General Population

Household surveys are moving from interviewer-administered modes to self-administered modes for data collection, but many households do not accurately follow within-household selection procedures in mail surveys. In this article, we examine accuracy of within-household selection using an oldest adult/youngest adult method in web, mail, and mixed-mode surveys. The frame for this study comes from a telephone survey conducted with Nebraska residents in which the oldest adult/youngest adult method is used to select the initial respondent. One year later, these telephone participants are followed up using identical household selection methods. This article examines characteristics of people who followed the selection procedures compared to those who did not.

Social isolation, survey nonresponse, and nonresponse bias: An empirical evaluation using social network data within an organization

Survey researchers have long hypothesized that social isolation negatively affects the probability of survey participation and biases survey estimates. Previous research, however, has relied on proxy measures of isolation, such as being a marginalized group member within a population. We re-examine the relationship between social isolation and survey participation using direct measures of social isolation derived from social network data; specifically, instrumental research and expressive friendship connections among faculty within academic departments. Using a reconceptualization of social isolation, we find that social network isolation is negatively associated with unit response. Among women (a numerical minority group within the organization), we further find that social group isolation (i.e., lacking instrumental network connections to men, the majority group in the organization) is negatively associated with survey participation. Finally, we show that some survey estimates are systematically biased due to nonparticipation from socially isolated people.

Effects of Smiley Face Scales on Visual Processing of Satisfaction Questions in Web Surveys

Web surveys permit researchers to use graphic or symbolic elements alongside the text of response options to help respondents process the categories. Smiley faces are one example used to communicate positive and negative domains. How respondents visually process these smiley faces, including whether they detract from the question’s text, is understudied. We report the results of two eye-tracking experiments in which satisfaction questions were asked with and without smiley faces. Respondents to the questions with smiley faces spent less time reading the question stem and response option text than respondents to the questions without smiley faces, but the response distributions did not differ by version. We also find support that lower literacy respondents rely more on the smiley faces than higher literacy respondents.

Using a Calendar and Explanatory Instructions to Aid Within-household Selection in Mail Surveys

Although researchers can easily select probability samples of addresses using the U.S. Postal Service’s Delivery Sequence File, randomly selecting respondents within households for surveys remains challenging. Researchers often place within-household selection instructions, such as the next or last birthday methods, in survey cover letters to select respondents. Studies show that up to 30% of selections are inaccurate and samples tend not to represent key demographics. This article tests two design elements—a calendar and explanatory wording of selection instructions in cover letters—to aid and motivate households to carry out selection procedures accurately. We empirically examine these elements in two mail surveys of Nebraskans—the 2012 Nebraska Annual Social Indicators Survey and 2013 Nebraska Trees and Forest Survey. We find that neither the calendar nor the explanatory wording adversely affected response rates but that neither improved the representativeness of the completed samples and that the calendar actually reduced selection accuracy.