Glenn Deane

The Spatial Patterning of County Homicide Rates: An Application of Exploratory Spatial Data Analysis

The possibility that homicides can spread from one geographic area toanother has been entertained for some time by social scientists, yetsystematic efforts to demonstrate the existence, or estimate the strength,of such a diffusion process are just beginning. This paper uses exploratoryspatial data analysis (ESDA) to examine the distribution of homicides in 78counties in, or around, the St. Louis metropolitan area for two timeperiods: a period of relatively stable homicide (1984–1988) and aperiod of generally increasing homicide (1988–1993). The findingsreveal that homicides are distributed nonrandomly, suggestive of positivespatial autocorrelation. Moreover, changes over time in the distribution ofhomicides suggest the possible diffusion of lethal violence out of onecounty containing a medium-sized city (Macon County) into two nearbycounties (Morgan and Sangamon Counties) located to the west. Althoughtraditional correlates of homicide do not account for its nonrandom spatialdistribution across counties, we find some evidence that more affluentareas, or those more rural or agricultural areas, serve as barriers againstthe diffusion of homicides. The patterns of spatial distribution revealedthrough ESDA provide an empirical foundation for the specification ofmultivariate models which can provide formal tests for diffusion processes.

Model selection procedures in social research: Monte-Carlo simulation results

Model selection strategies play an important, if not explicit, role in quantitative research. The inferential properties of these strategies are largely unknown, therefore, there is little basis for recommending (or avoiding) any particular set of strategies. In this paper, we evaluate several commonly used model selection procedures [Bayesian information criterion (BIC), adjusted R 2, Mallows’ C p, Akaike information criteria (AIC), AICc, and stepwise regression] using Monte-Carlo simulation of model selection when the true data generating processes (DGP) are known. We find that the ability of these selection procedures to include important variables and exclude irrelevant variables increases with the size of the sample and decreases with the amount of noise in the model. None of the model selection procedures do well in small samples, even when the true DGP is largely deterministic; thus, data mining in small samples should be avoided entirely. Instead, the implicit uncertainty in model specification should be explicitly discussed. In large samples, BIC is better than the other procedures at correctly identifying most of the generating processes we simulated, and stepwise does almost as well. In the absence of strong theory, both BIC and stepwise appear to be reasonable model selection strategies in large samples. Under the conditions simulated, adjusted R 2, Mallows’ C p AIC, and AICc are clearly inferior and should be avoided.

Analysis of PCB congeners related to cognitive functioning in adolescents.

To investigate the characteristics of PCBs that are linked to cognitive functioning, those congeners that were concurrently found in 271 Mohawk adolescents were grouped according to structure (dioxin-like or non-dioxin-like) and persistence (persistent or low-persistent). After the effects of the congener groups were orthogonalized, regression analyses (controlling for a number of variables found to be related to the cognitive outcomes) examined the relationship of each congener group to scores on three cognitive tests (the non-verbal Ravens Progressive Matrices, the Test of Memory and Learning, and the Woodcock Johnson-Revised). Five subtests from these cognitive tests were found to be associated with one or more PCB congener groups, most often at a moderate level. Two measures of long-term memory (Delayed Recall and Long Term Retrieval) were associated with all four congener groups. Nevertheless, examination of the role of individual congeners in the significantly related congener groups revealed that almost all congeners associated with cognitive outcomes were non-dioxin-like and ortho-substituted. A notable exception was the Ravens test where scores were associated only with dioxin-like congeners. This finding adds to the limited evidence of neurotoxic effects of dioxin-like congeners. Auditory Processing was related only to the persistent congener group. The association of the non-persistent congener group with three cognitive test scores (Delayed Recall, Long Term Retrieval and Comprehension-Knowledge) suggests that the Mohawk adolescents have experienced continuing or recent environmental exposure to PCBs that is sufficient to result in detectable cognitive decrements. Comparison of our findings with those of other human studies was limited by the relative lack of specificity of both PCB measures and cognitive outcome measures in much previous work.

The Role of Immigrant Enclaves for Latino Residential Inequalities

We investigate the difference that immigrant enclaves make for the residential contexts of Latino families in the USA. We argue that enclaves may no longer function simply as temporary way stations, the classic depiction of them, because of the compromised legal status of many Latinos. We examine this role with an innovative method that uses publicly available census tabulations (from the 2000 Census, in our case) to develop HLM models, in which race/ethnicity and income are controlled at the family level, along with neighbourhood context and metropolitan characteristics. Comparing Latino residential patterns to those of whites and blacks reveals the large neighbourhood disadvantages of Latinos, which except for greater exposure to whites are on the order of those suffered by African-Americans. We find that Hispanic families improve their residential situations as their incomes go up and, usually, also when they live in suburbs. But residence outside of immigrant enclaves produces the largest positive changes. The enclaves are a fundamentally different kind of residential space in which the potential for neighbourhood improvement is modest.

Life-Course Changes and Parent–Adult Child Contact

Despite increased interest in parent–adult child relations, there has been little attention to how these are influenced by changes in their lives, reflecting transitions and linked lives within a life-course perspective. Hybrid multilevel models are used to analyze the change in parent–adult child contact over two waves of the National Survey of Families and Households. Changes in parent–child proximity, parent and child marital status, and child parental status are associated with change in contact; continued coresidence with another adult child is related to contact with non-coresidential children; but change in parent health does not affect contact. Some patterns are stronger for daughters and biological children who tend to have stronger relationships with their parents. These analyses demonstrate how life-course transitions of parents and adult children can be examined in family context to understand how changes in the life of one family member may influence relations with another.

Relationships of Polychlorinated Biphenyls and Dichlorodiphenyldichloroethylene (p,p’-DDE) with Testosterone Levels in Adolescent Males

Background: Concern persists over endocrine-disrupting effects of persistent organic pollutants (POPs) on human growth and sexual maturation. Potential effects of toxicant exposures on testosterone levels during puberty are not well characterized. Objectives: In this study we evaluated the relationship between toxicants [polychlorinated biphenyls (PCBs), dichlorodiphenyldichloroethylene (p,p´-DDE), hexachlorobenzene (HCB), and lead] and testosterone levels among 127 Akwesasne Mohawk males 10 to < 17 years of age with documented toxicant exposures. Methods: Data were collected between February 1996 and January 2000. Fasting blood specimens were collected before breakfast by trained Akwesasne Mohawk staff. Multivariable regression models were used to estimates associations between toxicants and serum testosterone, adjusted for other toxicants, Tanner stage, and potential confounders. Results: The sum of 16 PCB congeners (Σ16PCBs) that were detected in ≥ 50% of the population was significantly and negatively associated with serum testosterone levels, such that a 10% change in exposure was associated with a 5.6% decrease in testosterone (95% CI: –10.8, –0.5%). Of the 16 congeners, the more persistent ones (Σ8PerPCBs) were related to testosterone, whereas the less persistent ones, possibly reflecting more recent exposure, were not. When PCB congeners were subgrouped, the association was significant for the sum of eight more persistent PCBs (5.7% decrease; 95% CI: –11, –0.4%), and stronger than the sum of six less persistent congeners (3.1% decrease; 95% CI: –7.2, 0.9%). p,p´-DDE was positively but not significantly associated with serum testosterone (5.2% increase with a 10% increase in exposure; 95% CI: –0.5, 10.9%). Neither lead nor HCB was significantly associated with testosterone levels. Conclusions: Exposure to PCBs, particularly the more highly persistent congeners, may negatively influence testosterone levels among adolescent males. The positive relationship between p,p´-DDE and testosterone indicates that not all POPs act similarly. Citation: Schell LM, Gallo MV, Deane GD, Nelder KR, DeCaprio AP, Jacobs A; Akwesasne Task Force on the Environment. 2014. Relationships of polychlorinated biphenyls and dichlorodiphenyldichloroethylene (p,p´-DDE) with testosterone levels in adolescent males. Environ Health Perspect 122:304–309; http://dx.doi.org/10.1289/ehp.1205984

Cross-Sibling Effects in Parent-Adult Child Exchanges of Socioemotional Support

The authors use National Survey of Families and Households first wave data and innovative modeling to examine how one parent-adult child pair may affect other pairs. Three conceptual models guide the analyses of parents’ giving and receiving of socioemotional support, representing enhancement, compensation, and independence. Giving support to one child is related to more giving to others (enhancement), but receiving support from one child is related to less receipt from others (compensation). Cross-sibling interactions do not reveal significant effects of distance of one child on exchanges with others, nor of gender or stepchild status of adult children. Cross-sibling interactions differ by race, suggesting enhancement in receiving support among Blacks and enhancement in giving support among non-Blacks. These analyses demonstrate the value of examining how parent-adult child ties are influenced by each other and by their family context.

Navigating time and space in population studies

Preface.- Acknowledgements.- Introduction.- Chapter 1: An Innovative Methodology for Space-Time Analysis with an Application to the 1960-2000 Brazilian Mortality Transition: Carl P. Schmertmann, Joseph E. Potter, and Renato M. Assuncao.- Chapter 2: Spatial Aspects of the American Fertility Transition in the Nineteenth Century: Michael R. Haines and J. David Hacker.- Chapter 3: Spatial and Temporal Patterns of Fertility Transition in Muslim Populations: Hani A. Guend.- Chapter 4: Spatial and Temporal Analyses of Surname Distributions to Estimate Mobility and Changes in Historical Demography: The Example of Savoy (France) from the Eighteenth to the Twentieth Century: Pierre Darlu, Guy Brunet, and Dominique Barbero.- Chapter 5: Widening Horizons? The Geography of the Marriage Market in Nineteenth- and Early-Twentieth Century Netherlands: Peter Ekamper, Frans van Poppel, and Kees Mandemakers.- Chapter 6: Finding Frontiers in the U.S. Great Plains from the End of the Civil War to the Eve of the Great Depression: Myron P. Gutmann, Glenn D. Deane, and Kristine Witkowski.- Chapter 7: Commonalities and Contrasts in the Development of Major United States Urban Areas: A Spatial and Temporal Analysis from 1910 to 2000: Andrew A. Beveridge.- Chapter 8: Economic Transition and Social Inequality in Early Twentieth Century Puerto Rico: Katherine J. Curtis.- Index.

The Interchangeability of Homicide Data Sources: A Spatial Analytical Perspective

The authors extend the work of Wiersema, Loftin and McDowall (2000) by assessing the interchangeability of Supplemental Homicide Report (SHR) and National Center for Health Statistics (NCHS) homicide rates in the context of spatial analysis. Conclusions drawn from spatial analyses may be affected by the choice of data source if the underrecording of homicide tends to cluster in geographic space more strongly in one data source than the other. Such a situation could alter indications of homicide rate clustering and substantive conclusions drawn from multivariate spatial regression models. Results indicate that although these data sources are interchangeable for urban counties, spatial effects and the effects of social structural factors vary by data source for rural counties.

Finding Frontiers in the U.S. Great Plains from the End of the Civil War to the Eve of the Great Depression

This chapter uses a new approach to studying patterns of spatial settlement to understand the forces that shaped the movement of the European-origin population into the semi-arid and arid natural region referred to as the Great Plains of the United States, between 1880 and 1940. Defining settlement as the process by which each of the roughly 500 counties in the region reached a population density threshold of four persons per square mile, the results evaluate hypotheses that suggest that the most important forces at work were a combination of structural attributes of the national process (population moving from east to west), climate (precipitation, temperature), other resources inside the region (irrigation, transportation, energy, employment in industry) and developments outside the region, such as the need to supply food to gold and silver miners working in the mountainous region to the west. The approach taken in the chapter includes a new strategy for working with changing county boundaries and a statistical method employing Cox proportional hazards models for repeated events. The results reveal a process of settlement diffusion in the Great Plains and demonstrate that variations in that diffusion process favoured areas well-suited to cropping, mining and manufacturing.