Daniel C. Coster

ShopTalk: Independent Blind Shopping Through Verbal Route Directions and Barcode Scans

Independent shopping in modern grocery stores that carry thousands of products is a great challenge for people with visual impairments. ShopTalk is a proof-of-concept wearable system designed to assist visually impaired shoppers with finding shelved products in grocery stores. Using synthetic verbal route directions and descriptions of the store lay- out, ShopTalk leverages the everyday orientation and mobility skills of independent visually impaired travelers to direct them to aisles with target products. Inside aisles, an off-the-shelf barcode scanner is used in conjunction with a software data structure, called a barcode connectivity matrix, to locate target product on shelves. Two experiments were performed at a real world supermarket. A successful earlier single-subject experiment is summarized and a new experiment involving ten visually impaired participants is presented. In both experiments, ShopTalk was successfully used to guide visually im- paired shoppers to multiple products located in aisles on shelves. ShopTalk is a feasible system for guiding visually im- paired shoppers who are skilled, independent travelers. Its design does not require any hardware instrumentation of the store and leads to low installation and maintenance costs.

Interest in a Stepped Approach Model (SAM): Identification of Recruitment Strategies for University Alcohol Programs

This study evaluates interest in a stepped approach model (SAM) of service delivery and identifies variables to enhance recruitment to alcohol programs. Subjects were 2,443 college student drinkers (1,420 men and 1,023 women) at a large midwestern university (selected by stratified systematic sampling) who completed a questionnaire on drinking behavior and interest in five intervention steps. Results indicated that there was more interest in interventions that required less time, which supports predictions of SAM. Overall lack of interest peaked at 67% for men and 65% for women and did not change significantly across the last three steps of SAM, which included traditional interventions of group and individual counseling. The data suggest that the stepped approach is more viable than conventional approaches, recruitment is unquestionably an exigent research priority because of the overwhelming lack of interest in alcohol programs, and specific variables that were identified may be important for recruiting particular subgroups.

Evaluation of a School-based Teen Obesity Prevention Minimal Intervention

OBJECTIVE A school-based nutrition education minimal intervention (MI) was evaluated. DESIGN The design was experimental, with random assignment at the school level. SETTING Seven schools were randomly assigned as experimental, and 7 as delayed-treatment. PARTICIPANTS The experimental group included 551 teens, and the delayed treatment group included 329 teens. INTERVENTION The minimal intervention was Present and Prevent, a commercially available PowerPoint program presented in two 30-minute time slots over 1 week. MAIN OUTCOME MEASURES The dependent variables were nutrition knowledge, attitudes, peer and family influences, behavioral intentions, and program satisfaction. The independent variable was group assignment. ANALYSES A matched-pairs and 2-sample t test were used respectively to assess within-group and between-group changes. RESULTS Significant experimental posttest improvements occurred in the following: knowledge (P < .001); intention to maintain a healthy body weight because of importance to friends (P < .001); and intention to eat fewer fried foods, eat fewer sweets, look more at food labels, and limit TV watching (all P < .001). Program satisfaction measures were significantly associated with each of the healthy weight maintenance behavioral intentions. CONCLUSIONS AND IMPLICATIONS The MI teen obesity prevention program made an impact on nutrition knowledge and positive behavioral intentions in only 2 classroom sessions and was well received by participants.

Monitoring wildlife-vehicle collisions in the information age: how smartphones can improve data collection.

Background Currently there is a critical need for accurate and standardized wildlife-vehicle collision data, because it is the underpinning of mitigation projects that protect both drivers and wildlife. Gathering data can be challenging because wildlife-vehicle collisions occur over broad areas, during all seasons of the year, and in large numbers. Collecting data of this magnitude requires an efficient data collection system. Presently there is no widely adopted system that is both efficient and accurate. Methodology/Principal Findings Our objective was to develop and test an integrated smartphone-based system for reporting wildlife-vehicle collision data. The WVC Reporter system we developed consisted of a mobile web application for data collection, a database for centralized storage of data, and a desktop web application for viewing data. The smartphones that we tested for use with the application produced accurate locations (median error = 4.6–5.2 m), and reduced location error 99% versus reporting only the highway/marker. Additionally, mean times for data entry using the mobile web application (22.0–26.5 s) were substantially shorter than using the pen/paper method (52 s). We also found the pen/paper method had a data entry error rate of 10% and those errors were virtually eliminated using the mobile web application. During the first year of use, 6,822 animal carcasses were reported using WVC Reporter. The desktop web application improved access to WVC data and allowed users to easily visualize wildlife-vehicle collision patterns at multiple scales. Conclusions/Significance The WVC Reporter integrated several modern technologies into a seamless method for collecting, managing, and using WVC data. As a result, the system increased efficiency in reporting, improved accuracy, and enhanced visualization of data. The development costs for the system were minor relative to the potential benefits of having spatially accurate and temporally current wildlife-vehicle collision data.

Online Physical Activity Information: Will Typical Users Find Quality Information?

This study evaluated physical activity Web sites to determine quality, accuracy, and consistency with principles of the extended parallel process model (EPPM). Three keyword searches were conducted using 4 search engines to find a sample of N = 41 Web sites. Three raters evaluated the Web sites using the JAMA benchmarks to assess quality and American College of Sports Medicine and Centers for Disease Control and Prevention recommendations for physical activity to determine accuracy, as well as checking for inclusion of EPPM variables. Data were analyzed using descriptive statistics and analysis of variance with least squares means. Only 22% of the sites were high quality, none were highly accurate, and most were consistent with the EPPM. Quality ratings were weakly associated with accuracy. Educational and .net sites were rated significantly higher in quality and accuracy, and government sites were most consistent with the EPPM. Quality Web sites were more often found by using Yahoo and Google. “Exercise” yielded more accurate results, whereas “physical activity” and “fitness” produced more Web sites consistent with the EPPM. It is encouraging that most sites incorporated EPPM concepts; however, quality and accuracy were poor, leaving physical activity information seekers at risk for disease and injury.

Adjusting the Census of 1990 The Smoothing Model

Considering the difficulties, the Census Bureau does a remarkably good job at counting people. This article discusses techniques for adjusting the census. If there is a large undercount, these techniques may be accurate enough for adjustment. With a small undercount, they are unlikely to improve on the census; instead, adjustment could easily degrade the accuracy of the data. The focus will be sampling error, that is, uncertainty in estimates due to the luck of the draw in choosing the sample. Sampling error is a major obstacle to adjusting the 1990 census, even at the state level. To control sampling error, the Census Bureau used a smoothing model. However, the model does not solve the problem, because its effects are strongly dependent on unverified and implausible assumptions. This story has a broader moral. Statistical models are often defended on grounds of robustness, that is, estimates do not depend strongly on assumptions. But the standard errors, which are internally generated measures of precision, may be critical. Then caution is in order. If the model is at all complicated, the standard errors may turn out to be driven by assumptions not data—the antithesis of robustness.

Tables of minimum cost, linear trend-free run sequences for two- and three-level fractional factorial designs

Abstract For the two- and three-level fractional factorial designs tabled in National Bureau of Standards Applied Mathematics Series publications 48 and 54, plus the resolution V plans developed by Addelman ( Technometrics 7, 1965), we present generator sequences that may be used with the Generalized Foldover Scheme of Coster and Cheng ( Ann. Statist. 16 , 1988) to produce systematic run orders which minimize, or nearly minimize, a cost function equal to the number of times the factors change levels during the time sequence in which the runs are performed and which simultaneously have all factor main effects components orthogonal to a polynomial time trend.

How does variation in winter weather affect deer—vehicle collision rates?

Understanding how deer move in relationship to roads is critical, because deer are in vehicle collisions, and collisions cause vehicle damage, as well as human injuries and fatalities. In temperate climates, mule deer Odocoileus hemionus have distinct movement patterns that affect their spatial distribution in relationship to roads. In this paper, we analyzed deer movements during two consecutive winter seasons with vastly different conditions to determine how deer—vehicle collision rates responded. We predicted that deer—vehicle collision rates would be higher when precipitation and snow depth were higher. We used meteorological data from local weather stations to describe temperature, precipitation and snow depth. We monitored deer movements with global positioning system telemetry to document distance of deer to roads, elevation use and road crossing rates. We also documented changes in deer abundance and traffic volumes, which were potentially confounding variables. We found that precipitation decreased 50% and snow depth decreased 48% between winters. In response, deer used habitats that were 16% higher in elevation and that were 213% farther from roads with high traffic volumes. Consequently, crossing rates also decreased as much as 96% on roads with high traffic volumes. Reduced crossing rates were likely responsible for much of the 75% decrease in deer—vehicle collisions that occurred during the second winter. Abundance and traffic volume also can be important factors affecting deer—vehicle collisions rates. However, it is unlikely they were the major drivers of variation in deer—vehicle collisions during our study, because traffic volumes did not change between years and deer abundance only decreased 7%. Our data suggest a mechanism by which variation in winter conditions can contribute to differences in deer—vehicle collision rates between years. These findings have significant management implications for deer—vehicle collision mitigation.

Effects of saline drinking water on early gosling development

Relatively high levels of saline drinking water may adversely affect the growth, development, and survival of young waterfowl. Saline drinking water was suspect in the low survival rate of Canada goose (Branta canadensis) goslings at Fish Springs National Wildlife Refuge (FSNWR) in western Utah. Hence, we investigated the effects of saline drinking water on the survival and growth of captive, wild-strain goslings from day 1-28 following hatch. We compared survival and growth (as measured by body mass, wing length, and culmen length) between a control group on tap water with a mean specific conductivity of 650 μS/cm, and 2 saline water treatments: (1) intermediate level (12,000 μS/cm), and (2) high level (18,000 μS/cm). Gosling mortality occurred only in the 18,000 μS/c treatment group (33%; n = 9). Slopes of regressions of mean body mass, wing length, and culmen length on age were different from each other (P < 0.05), except for culmen length for the intermediate and high treatment levels. We predict that free-ranging wild goslings will experience mortality at even lower salinity levels than captive goslings because of the combined effects of depressed growth and environmental stresses, including hot desert temperatures and variable food quality over summer.

Shoptalk: toward independent shopping by people with visual impairments

ShopTalk, a proof-of-concept system designed to assist individuals with visual impairments with finding shelved products in grocery stores, is built on the assumption that simple verbal route directions and layout descriptions can be used to leverage the O&M skills of independent visually impaired travelers to enable them to navigate the store and retrieve shelved products. This paper introduces ShopTalk and summarizes experiments performed in a real-world supermarket.