個體選擇模式(discrete choice model)廣泛應用於國外的交通運輸及行銷領域，而國內交通運輸領域，也長期以此模式分析個體的運具選擇行為。反觀國內的行銷領域，因較難取得消費者的商品品牌購買紀錄，而鮮少應用個體選擇模式分析消費者的選擇行為。有鑒於此，本研究嘗試以問卷收集消費者對三個洗髮精品牌的選擇行為，以個體選擇模式中的多項邏輯模式(multinomial logit model)、巢狀邏輯模式(nested multinomial logit model)、混合多項邏輯模式(mixed logit model)進行分析，檢驗問卷設計中的促銷活動、消費者特性對選擇行為的影響性。 實證分析的結果發現，洗髮精的原價格及促銷折扣、贈品容量、加量不加價等促銷活動，皆對消費者的選擇行為有顯著的影響力，其中促銷折扣與贈品容量影響的程度較大，是較具有效果的促銷活動。而消費者的性別、年齡、職業及品牌更換的頻率，皆影響洗髮精的選擇行為。此外，消費者若固定選擇自己最常購買的洗髮精，此類型的消費者與其他人的品牌選擇行為，也有顯著的不同。 此外，根據本研究樣本，我們也發現海倫仙度絲與潘婷間的替代、互補性較強。 / Discrete choice model has been demonstrated to be a useful tool for analyzing consumers’ choice behavior data in the area of transportation and marketing research. However, since a complete data set containing consumers’ history of purchase behavior was rarely available to public, the model was less popular in the marketing research area than in the transportation research in Taiwan. Based on limited survey data on consumers’ choice among three different brands of shampoo, we applied multinomial logit model、nested multinomial logit model、mixed logit model in this study to understand promotion program’s effect on consumers’ choice behavior , the result showed that shampoos’ original price、discount、volume of hair conditioner bestowal、more volume with the same price all had significant impacts on consumers’ choice behavior, among them, discount and volume of hair conditioner bestowel influenced more .In addition, consumers’ gender、age、occupation and frequency of changing brands also affected consumers on choosing brands of shampoos. The study also found that a consumer who chose the same brand regularly behaved notably differently.

Travel demand Management (TDM) focuses on improving the efficiency of the transportation system through changing traveler's travel behavior rather than expanding the infrastructure. An employer based integrated TDM program generally includes strategies designed to change the commuter's travel behavior in terms of mode choice, time choice and travel frequency. Research on TDM has focused on the evaluation of the effectiveness of TDM program to report progress and find effective strategies. Another research area, identified as high-priority research need by TRB TDM innovation and research symposium 1994 [Transportation Research Circular, 1994], is to develop tools to predict the impact of TDM strategies in the future. These tools are necessary for integrating TDM into the transportation planning process and developing realistic expectations. Most previous research on TDM impact evaluation was worksite-based, retrospective, and focused on only one or more aspects of TDM strategies. That research is generally based on survey data with small sample size due to lack of detailed information on TDM programs and promotions and commuter travel behavior patterns, which cast doubts on its findings because of potential small sample bias and self-selection bias. Additionally, the worksite-based approach has several limitations that affect the accuracy and application of analysis results. Based on the Washington State Commute Trip Reduction (CTR) dataset, this dissertation focuses on analyzing the participation rates of compressed work week schedules and telecommuting for the CTR affected employees, modeling the determinants of commuter's compressed work week schedules and telecommuting choices, and analyzing the quantitative impacts of an integrated TDM program on individual commuter's mode choice. The major findings of this dissertation may have important policy implications and help TDM practitioners better understand the effectiveness of the TDM strategies in terms of person trip and vehicle trip reduction. The models developed in this dissertation may be used to evaluate the impacts of an existing TDM program. More importantly, they may be incorporated into the regional transportation model to reflect the TDM impacts in the transportation planning process.

This study uses conjoint analysis to examine the preferences of buyers for Bureau of Land Management (BLM) wild horses based on physical attributes of wild horses and individual characteristics of the buyers. Generalized ordered logit models and multinomial logit models are used to study the impact of the buyers’ demographic characteristics such as age, gender, knowledge about wild horse care, and number of wild horses previously adopted on physical attributes of the horses such as color, age, height, training status, temperament, conformation, and unique markings. Using a choice experiment, taken together, these attributes determine buyer’s preferences for a wild horse. This study reveals that characteristics of buyers have significant effects on their preferences for wild horses. Their gender, age, knowledge about wild horse care, and the number of horses previously adopted influence the importance that buyers place on physical attributes of a wild horse in their decision to purchase a wild horse.

Artificial Intelligence in form of Machine Learning classifiers is increasingly applied for travel choice modeling issues and therefore constitutes a promising, competitive alternative towards conventional discrete choice models like the Logit approach. In comparison to traditional theory-based models, data-driven Machine Learning generally shows powerful predictive performance, but often lacks in model interpretability, i.e., the provision of comprehensible explanations of individual decision behavior. Consequently, the question about which approach is superior remains unanswered. Thus, this paper performs an in-depth comparison between benchmark Logit models and Artificial Neural Networks and Decision Trees representing two popular algorithms of Artificial Intelligence. The primary focus of the analysis is on the models’ prediction performance and its ability to provide reasonable economic behavioral information such as the value of travel time and demand elasticities. For this purpose, I use crossvalidation and extract behavioral indicators numerically from Machine Learning models by means of post-hoc sensitivity analysis. All models are specified and estimated on synthetic and empirical data. As the results show, Neural Networks provide plausible aggregate value of time and elasticity measures, even though their values are in different regions as those of the Logit models. The simple Classification Tree algorithm, however, appears unsuitable for the applied computation procedure of these indicators, although it provides reasonable interpretable decision rules for travel choice behavior. Consistent with the literature, both Machine Learning methods achieve strong overall predictive performance and therefore outperform the Logit models in this regard. Finally, there is no clear indication of which approach is superior. Rather, there seems to be a methodological tradeoff between Artificial Intelligence and discrete choice models depending on the underlying modeling objective.

Masters of Science / Department of Agricultural Economics / Hikaru H. Peterson / Concerns about environmental degradation due to agriculture have gained importance as it is associated with soil erosion, health hazards, and ground water pollution. Environment-friendly land use practices have been developed to gain a wide range of environmental benefits including reduced soil erosion, reduced nutrient runoff from crop and livestock facilities, increased biodiversity preservation efforts, and restoration of wetlands and other native ecosystems. No-till is one such practice where soil erosion, nutrient runoff and environmental degradation can be reduced to a certain extent. This study evaluated the factors affecting the adoption of tillage systems in Kansas. A survey was conducted with a total of 135 participants from four different locations in the state of Kansas between August 2006 and January 2007. The adoption process was modeled as a two-step econometric models consisting of perception and adoption equations to estimate the impacts of demographic variables and farmers’ familiarity with and participation in certain conservation programs. The results for the perception models showed that the farm operators’ perceptions regarding whether BPM installation and management is unfair to producers or not and whether environmental legislation is often unfair to producers do not vary systematically across farm size, producers’ familiarity and participation in conservation programs, or other demographics considered in the study. On the other hand, their perceptions regarding how polluted their water supplies varied by their thoughts on relative profitability across various tillage practices, their primary occupation, and their familiarity with conservation programs. Specifically, the results suggested that those who regarded no-till practices to be more profitable than other tillage practices or whose primary occupation was farming-related tended to believe that ground water was not polluted, and those who were less familiar with available conservation programs tended to believe that surface waters were not polluted. The adoption model results suggested that farmers with greater operating acreage, those who perceived that no-till was more profitable than other tillage systems, and those with greater familiarity with and participation in existing conservation programs were more likely to adopt more conservation tillage systems, all else equal. Further, perceptions of fairness of environmental regulations or the level of pollution did not impact the tillage choices.

The study aims to identify the influence of socioeconomic factors on the prevalence of type 2 diabetes for individuals aged 27 and older in the Republic of Belarus. We analyze data from the Diabetes Survey conducted by the Endocrinology Medical Center in Minsk and the Ministry of Health of the Republic of Belarus from 2011 to 2015. The association between socioeconomic factors and the prevalence of type 2 diabetes is examined using logistic regression with sequential adjustments for clinical and behavioral predictors. Our findings indicate that individuals with lower income and educational levels are more likely to suffer from type 2 diabetes than those in higher income and education groups. Moreover, the prevalence of type 2 diabetes decreases as income and educational level go up. Furthermore, this association remains significant even after further adjusting for various behavioral and clinical factors. In addition, we confirm that type 2 diabetes is more prevalent among overweight / obese, physically inactive and older individuals. These findings suggest that strategies for preventive diabetes programs should be focused on socioeconomic environment rather than on individual risky behavior only.

Master of Science / Department of Civil Engineering / Sunanda Dissanayake / In order to improve safety of the overall surface transportation system, each of the critical areas needs to be addressed separately with more focused attention. Statistics clearly show that large-truck crashes contribute significantly to an increased percentage of high-severity crashes. It is therefore important for the highway safety community to identify characteristics and contributory causes related to large-truck crashes. During the first phase of this study, fatal crash data from the Fatality Analysis Reporting System (FARS) database were studied to achieve that objective. In this second phase, truck-crashes of all severity levels were analyzed with the intention of understanding characteristics and contributory causes, and identifying factors contributing to increased severity of truck-crashes, which could not be achieved by analyzing fatal crashes alone. Various statistical methodologies such as cross-classification analysis and severity models were developed using Kansas crash data. Various driver-, road-, environment- and vehicle- related characteristics were identified and contributory causes were analyzed. From the cross-classification analysis, severity of truck-crashes was found to be related with variables such as road surface (type, character and condition), accident class, collision type, driver- and environment-related contributory causes, traffic-control type, truck-maneuver, crash location, speed limit, light and weather conditions, time of day, functional class, lane class, and Average Annual Daily Traffic (AADT). Other variables such as age of truck driver, day of the week, gender of truck-driver, pedestrian- and truck-related contributory causes were found to have no relationship with crash severity of large trucks. Furthermore, driver-related contributory causes were found to be more common than any other type of contributory cause for the occurrence of truck-crashes. Failing to give time and attention, being too fast for existing conditions, and failing to yield right of way were the most dominant truck-driver-related contributory causes, among many others. Through the severity modeling, factors such as truck-driver-related contributory cause, accident class, manner of collision, truck-driver under the influence of alcohol, truck maneuver, traffic control device, surface condition, truck-driver being too fast for existing conditions, truck-driver being trapped, damage to the truck, light conditions, etc. were found to be significantly related with increased severity of truck-crashes. Truck-driver being trapped had the highest odds of contributing to a more severe crash with a value of 82.81 followed by the collision resulting in damage to the truck, which had 3.05 times higher odds of increasing the severity of truck-crashes. Truck-driver under the influence of alcohol had 2.66 times higher odds of contributing to a more severe crash. Besides traditional practices like providing adequate traffic signs, ensuring proper lane markings, provision of rumble strips and elevated medians, use of technology to develop and implement intelligent countermeasures were recommended. These include Automated Truck Rollover Warning System to mitigate truck-crashes involving rollovers, Lane Drift Warning Systems (LDWS) to prevent run-off-road collisions, Speed Limiters (SLs) to control the speed of the truck, connecting vehicle technologies like Vehicle-to-Vehicle (V2V) integration system to prevent head-on collisions etc., among many others. Proper development and implementation of these countermeasures in a cost effective manner will help mitigate the number and severity of truck-crashes, thereby improving the overall safety of the transportation system.

Este trabalho procurou avaliar as correlações existentes entre as características dos docentes, da escola, dos diretores, além da percepção dos professores e do diretor, e da conjuntura econômica sobre a decisão do professor de abandonar a rede pública estadual paulista. Os resultados indicam que as características dos docentes exercem uma forte influência sobre a decisão destes de evadir. Em relação às características da escola, tanto para o PEB I quanto para o PEB II o número de matrículas foi estatisticamente significantes para explicar a evasão, assim como, a experiência do diretor. Apenas para o PEB II o resultado da escola no Saresp foi relevante, a característica familiar dos alunos diferiu no seu comportamento entre os modelos. As variáveis de percepção também apresentaram significância em alguns dos quesitos analisados. Desta forma, conclui-se, neste estudo, que diferentes aspectos intervêm na decisão do docente de permanecer na rede e, ainda que com alguns resultados contraditórios, o ambiente de trabalho parece ser um fator relevante nesta decisão / This study aimed to evaluate the correlation between the characteristics of teachers, school, directors, the perception of teachers and the director, and the economic environment on the decision of the teacher to abandon the São Paulo State public schools. The results indicate that the characteristics of teachers has a strong influence on the dropout decision. In relation to the characteristics of the school, its enrollment number and principal experience were statistically significant to explain the teacher\'s dropout. Only for those teachers who attends students from 5th grade or more, the performance of students in the SARESP test was important. In addition the family background was also important, but with different results according to types of teachers. In some dimensions, the perception of the Director and of the teachers also was significant. So we conclude that different variables influence the dropout decision of teachers. The characteristics of the teachers itself, as well as characteristics of the school can explain why teachers get out their Jobs in the São Paulo state schools

Um dos desafios da educação brasileira é a grande quantidade de jovens que deveriam estar matriculados no Ensino Médio, mas que não estão frequentando a escola. Além disso, a taxa de abandono escolar nos anos iniciais do ensino médio é muito superior àquela encontrada no último ano do ensino fundamental. Nesse sentido, esse trabalho se propõe a investigar qual o papel da família no processo de transição e permanência no ensino médio. Para tanto, devido a disponibilidade de dados, será investigada uma coorte de alunos aprovados em 2010 no 9° ano do Ensino Fundamental no Estado do Ceará. Será estimado um sequential logit model, cujos regressores serão, além de algumas características individuais, informações referentes ao status socioeconômico da família e ao ambiente familiar. Existe uma vasta literatura que vem se desenvolvendo desde Mare (1980) buscando compreender o papel do background familiar na desigualdade educacional, tratando o processo de escolarização como uma sequência de decisões. Os resultados aqui encontrados, além de dialogar com essa literatura, apontam que a família é determinante, tanto para a entrada, como para a permanência no ensino médio, principalmente por meio da escolaridade dos pais. Esse efeito, todavia, é maior para aqueles que se defrontam com a decisão de entrada no ensino médio fora da idade ideal ou que exercem atividade remunerada quando não estão na escola. / One of the challenges of Brazilian education is the large amount of young people who should be enrolled in high school, but who are not attending school. In addition, the drop-out rate in the early high school years is much higher than that found in the last grade of elementary school. Thus, this study aims to investigate the role of the family in transition and permanence in high school. Therefore, due to data availability, a cohort of students approved in 2010 in the last grade of elementary school in the state of Ceará will be investigated. Will be estimated to sequential logit model, whose covariates are, individual characteristics, information regarding socioeconomic status of the family and the family environment. There is a vast literature, that has been developing since Mare (1980), trying to understand the role of family background on educational inequality, treating the educational process as a sequence of decisions. The results found here, as well as dialogue with the literature, show that the family is crucial both for entry and for staying in high school, mainly through parental education. This effect, however, is higher for those who are faced with the decision to enter high school outside the ideal age or to engage in paid work when they are not in school.

Work zone tends to cause hazardous conditions for drivers and construction workers since work zones generate conflicts between construction activities and the traffic, therefore aggravate the existing traffic conditions and result in severe traffic safety and operational problems. To address the influence of various factors on the crash severity is beneficial to understand the characteristics of work zone crashes. The understanding can be used to select proper countermeasures for reducing the crash severity at work zones and improving work zone safety. In this dissertation, crash severity models were developed to explore the factor impacts on crash severity for two work zone crash datasets (overall crashes and rear-end crashes). Partial proportional odds logistic regression, which has less restriction to the parallel regression assumption and provides more reasonable interpretations of the coefficients, was used to estimate the models. The factor impacts were summarized to indicate which factors are more likely to increase work zone crash severity or which factors tends to reduce the severity. Because the speed variety is an important factor causing accidents at work zone area, the work zone speed profile was analyzed and modeled to predict the distribution of speed along the distance to the starting point of lane closures. A new learning machine algorithm, support vector regression (SVR), was utilized to develop the speed profile model for freeway work zone sections under various scenarios since its excellent generalization ability. A simulation-based experiment was designed for producing the speed data (output data) and scenario data (input data). Based on these data, the speed profile model was trained and validated. The speed profile model can be used as a reference for designing appropriate traffic control countermeasures to improve the work zone safety.