Developing sound or reliable statistical models for analyzing vehicle crashes is very important in highway safety studies. A difficulty arises when crash data exhibit overdispersion. Over-dispersion caused by unobserved heterogeneity is a serious problem and has been addressed in a variety ways within the negative binomial (NB) modeling framework. However, the true factors that affect heterogeneity are often unknown to researchers, and failure to accommodate such heterogeneity in the model can undermine the validity of the empirical results. Given the limitations of the NB regression model for addressing over-dispersion of crash data due to heterogeneity, this research examined an alternative model formulation that could be used for capturing heterogeneity through the use of finite mixture regression models. A Finite mixture of Poisson or NB regression models is especially useful when the count data were generated from a heterogeneous population. To evaluate these models, Poisson and NB mixture models were estimated using both simulated and empirical crash datasets, and the results were compared to those from a single NB regression model. For model parameter estimation, a Bayesian approach was adopted, since it provides much richer inference than the maximum likelihood approach. Using simulated datasets, it was shown that the single NB model is biased if the underlying cause of heterogeneity is due to the existence of multiple counting processes. The implications could be poor prediction performance and poor interpretation. Using two empirical datasets, the results demonstrated that a two-component finite mixture of NB regression models (FMNB-2) was quite enough to characterize the uncertainty about the crash occurrence, and it provided more opportunities for interpretation of the dataset which are not available from the standard NB model. Based on the models from the empirical dataset (i.e., FMNB-2 and NB models), their relative performances were also examined in terms of hotspot identification and accident modification factors. Finally, using a simulation study, bias properties of the posterior summary statistics for dispersion parameters in FMNB-2 model were characterized, and the guidelines on the choice of priors and the summary statistics to use were presented for different sample sizes and sample-mean values.

Obesity, as defined as BMI at or above the 95th percentile on the Centers for Disease Control and Prevention’s growth charts, has increased almost 3-fold among children in the United States since 1980. Overweight in adolescence has been associated with increased fat retention and high blood pressure in adulthood, among other symptoms of metabolic syndrome. However, normative patterns of weight change in childhood have not been developed. Groups of children may follow different trajectory patterns of BMI change over time. If common trajectory patterns could be identified, and their risk factors and outcomes understood, more nuanced intervention with families and children at risk for obesity could be developed. This study used a national dataset of 1,364 children whose weight and length was measured 12 times from birth through 15 ½ years. Testing both latent class growth analysis and growth mixture modeling identified four distinct subgroups, or classes, of BMI growth trajectory from 24 months – 8th grade. These classes were compared on numerous demographic, biological, and psychosocial risk factors identified in previous research as related to obesity. Classes were differentiated primarily on the child’s BMI at 15 months, the mother’s BMI at 15 months, birth weight for age, and percent increase in birth weight. Being male, Black, and lower SES were also related to membership in the higher-BMI trajectory classes. Of the psychosocial factors, maternal sensitivity, maternal depression, and attachment classification were also related to BMI class. Membership in these trajectories strongly predicted weight-related and blood-pressure outcomes at 15 ½ years over and above individual risk factors, demonstrating that patterns of change themselves are highly influential. The best-fitting models of weight-related outcomes at 15 ½ years included change trajectory in combination with biological, psychosocial, and SES risk factors from 0-24 months, with R² ranging from .31 = .50. Characteristics predicting adolescent overweight can be identified in the first years of life and should trigger the development and implementation of early intervention protocols in obstetrics and pediatrics. / text

This research analyzes welfare conditions of Afro-Colombians vis-à-vis non Afro-Colombians using Amartya Sen's Capability Approach as the theoretical framework, and the latent variable modeling as the empirical method. Multiple Indicators Multiple Causes (MIMIC) models are estimated using data from the Colombian Quality of Life Survey, 2003. Two latent constructs, namely, 'knowledge' and 'being adequately sheltered', represent the two 'Capability' dimensions to be analyzed. Ethnic background appears to have a consistently negative influence after controlling statistically by a set of relevant variables (e.g. being poor, area, marital status, age and gender, among others) included in the models as exogenous "causes" or "determinants" of each capability dimension. This implies that the capability set or the freedom an Afro-descendant enjoys in achieving the life he or she wants in terms of 'knowledge' and 'shelter' is consistently lower than that of a non Afro-descendant (whites and mestizos). As a consequence, achieved welfare or functioning achievement as expressed in terms of aspects such as years of education or dwelling conditions in the household is and would be consistently lower for individual members of this ethnic group. This evidence points toward the proposition that embedded patterns of racial discrimination are limiting Afro-Colombian capabilities and individual agency, beyond income levels or even access to educational resources. Hence, from a capability perspective removing racial discrimination must be an explicit objective of developmental policy. Accordingly, national policy must be directed not only to improving access for Afro-Colombians to resources and economic wellbeing, as traditional analysis of class disparity argues, but also toward the nurturing and expansion of the real freedom they have to pursue the goals they value. Thus, development policy in Colombia must altogether work toward the improvement of resource access for Afro-descendants and toward the creation of specific mechanisms to enforce the judicial instruments to fight against racial discrimination. These laws and judicial mechanisms were created to open spaces for political, social and economic participation for this population group on an equal basis, as their fellow citizens of non African descent, and are yet to be fulfilled.

The conceptualization of intimacy within adolescent romantic relationships has typically taken a linear approach: Adolescents experience initial romantic encounters within a group context and progress towards an exclusive dyadic dating relationship. This study uses a person-centered approach and conceptualizes adolescent romance as multi-dimensional.In Study 1, a large, nationally representative dataset (the National Longitudinal Study of Adolescent Health) was used to classify 10th and 11th grade adolescents into ideal romantic relationship styles via Latent Class Analysis. Four classed emerged: Concealers (3.6%; n=276), Abstainers (32.6%; n=2508), Engagers (51.4% of the sample; n= 3955), and Family Builders (12.5%; n=959). Concealers, primarily non-White ethnicities, preferred low social/emotional involvement but moderate sexual activities. Most adolescents with same-sex attractions were concealers. Concealers reported the greatest miss-match between ideal and real relationship activities. Abstainers, predominantly females, preferred: high social/emotional activities, to talk less about contraception/STDs, and low sexual activities. Engagers, predominantly male and White, scored highest on all social, emotional, and physical activities (exception of 'seeing less of friends', 'sex', 'pregnancy', and 'marriage'). Family builders, overly-represented by Latino, preferred high social, emotional, and physical dimensions including seeing less of friends, sexual intercourse, pregnancy, and marriage. Moderate discrepancies occurred between ideal and real activities.Study 2 was a focus group study of White and Latino adolescents (N=75) entering 10th through 12th grades. Using a symbolic interactionism theoretical framework, adolescents described four types of sexual relationships within their social subjective realities: Going-out, dating, friends with benefits, and hooking up. Going-out relationships, an exclusive and emotionally/physically close relationship, were the most easily described and the most intense and committed relationships. Dating relationships, however, were the most common type of sexual relationship and were less easily defined, partially due to the ambiguity of the relationship itself which is to 'get to know each other'. These relationships were somewhat exclusive and required less obligations. Friends with benefits (primarily physical relationships) and hooking up (single physical encounters) were casual relationships that required little to no commitment.Findings are interpreted via a developmental/feminist lens. Gender inequality and sexual double standards are potent forces that continue to shape adolescent's sexual behaviors, feelings, and experiences.

The study aims to estimate the ability of different grouping techniques on categorical response. We try to find out how well do they work? Do they really find clusters when clusters exist? We use Cancer Problems in Living Scales from the ACS as our categorical data variables and lung cancer survivors as our studying group. Five methods of cluster analysis are examined for their accuracy in clustering on both real CPILS dataset and simulated data. The methods include hierarchical cluster analysis (Ward's method), model-based clustering of raw data, model-based clustering of the factors scores from a maximum likelihood factor analysis, model-based clustering of the predicted scores from independent factor analysis, and the method of latent class clustering. The results from each of the five methods are then compared to actual classifications. The performance of model-based clustering on raw data is poorer than that of the other methods and the latent class clustering method is most appropriate for the specific categorical data examined. These results are discussed and recommendations are made regarding future directions for cluster analysis research.

This thesis was to determine the prevalence and distribution of paratuberculosis in the Ontario dairy sheep and dairy goat industries, identify potential risk factors for herds which tested positive, evaluate the accuracy of seven commercially available individual and two bulk tank diagnostic tests in these two populations, and determine the circulating strains of Mycobacterium avium ssp. paratuberculosis in faecal isolates obtained. A cross-sectional study was conducted between October 2010 and August 2011 in 29 goat herds and 21 sheep flocks located in Ontario. On each farm, 20 lactating animals over the age of two years were randomly selected and faeces, blood, and milk were sampled from each animal, and a bulk milk sample from each herd. A questionnaire inquiring about herd management and biosecurity behaviours was also completed. The seven individual animal tests evaluated were: faecal culture using the BACTEC® MGIT™ 960 liquid culture system, direct faecal PCR (Tetracore®, Rockville, MD) based on the hspX gene, the Prionics® ELISA on serum and milk, the IDEXX® ELISA on serum and milk, and the agar gel immunodiffusion (AGID) test on serum. The test evaluations used both frequentist (faecal culture - reference test) and Latent Class Analysis/Bayesian (LCA/BM) methods (no reference test). In goat herds, faecal culture demonstrated the highest sensitivity (Se), 81.1% (LCA/BM). In sheep, while faecal culture demonstrated the highest Se, 49.5%, there was a small probability it was higher than faecal PCR Se at 42.4%. The bulk tank tests evaluated were the 'Hyper-ELISA' test and real-time PCR test based on IS900 (AntelBio®). While PCR did not demonstrate sufficiently high Se to be used as a herd-level test, the Hyper-ELISA performed well as a herd-level test identifying farms with high prevalence when the cut-off was reduced to 0.05. Overall herd-level apparent prevalence was 79.3% in goat herds and 57.1% in sheep flocks when faecal culture was the reference standard and true herd-level prevalence (LCA/BM) was 83.0% and 66.8% in each population respectively. This high prevalence reveals a need for the implementation of a small ruminant paratuberculosis control program in Ontario, Canada based on testing, improving youngstock management, and strengthening biosecurity practices. / AHSI, OMAF

No description available.

This paper presents the fully Bayesian analysis of the latent class model using a new approach towards MCMC estimation in the context of mixture models. The approach starts with estimating unidentified models for various numbers of classes. Exact Bayes' factors are computed by the bridge sampling estimator to compare different models and select the number of classes. Estimation of the unidentified model is carried out using the random permutation sampler. From the unidentified model estimates for model parameters that are not class specific are derived. Then, the exploration of the MCMC output from the unconstrained model yields suitable identifiability constraints. Finally, the constrained version of the permutation sampler is used to estimate group specific parameters. Conjoint data from the Austrian mineral water market serve to illustrate the method. (author's abstract) / Series: Report Series SFB "Adaptive Information Systems and Modelling in Economics and Management Science"

Conjoint analysis is one of the most important tools to support product development, pricing and positioning decisions in management practice. For this purpose various models have been developed. It is widely accepted that models that take consumer heterogeneity into account, outperform aggregate models in terms of hold-out tasks. The aim of our study is to investigate empirically whether predictions of choice-based conjoint models which incorporate heterogeneity can successfully be generalized to a whole market. To date no studies exist that examine the real world performance of choice-based conjoint models by use of aggregate scanner panel data. Our analysis is based on four commercial choice-based conjoint pricing studies including a total of 43 stock keeping units (SKU) and the corresponding weekly scanning data for approximately two years. An aggregate model serves as a benchmark for the performance of two models that take heterogeneity into account, hierarchical Bayes and latent class. Our empirical analysis demonstrates that, in contrast to the performance using hold-out tasks, the real world performance of hierarchical Bayes and latent class is similar to the performance of the aggregate model. Our results indicate that heterogeneity cannot be generalized to a whole market and suggest that aggregate models are sufficient to predict market shares. (author's abstract) / Series: Report Series SFB "Adaptive Information Systems and Modelling in Economics and Management Science"

In many studies, choice based conjoint analysis is used to build a market simulator to develop marketing strategies; i.e., shares-of-preference are taken as market share forecasts. However, conjoint data are collected in interview situations, which may differ considerably from real shopping behavior. In this paper, we test the internal and external validity of four commercial choice based conjoint pricing studies including a total of 43 brands. We use conjoint and sales data to assess the relative performance of two modern approaches to estimate conjoint parameters: the segmentation based Latent Class model and the individual level Hierarchical Bayes approach. Our paper confirms previous results of the internal superiority of the Hierarchical Bayes approach. The main result of our investigation is that internal validity does not predict external validity and that Latent Class shows the same real world performance as Hierarchical Bayes. Both models show an average error of 4.2% in market share level prediction and a correlation of 69% between conjoint forecasts and real market shares. (author's abstract) / Series: Working Papers SFB "Adaptive Information Systems and Modelling in Economics and Management Science"