Apresenta-se um procedimento para estimativa da matriz origem-destino (O/D) a partir de um modelo de escolha discreta em combinação com uma matriz O/D de transporte público, a qual pode ser estimada de maneira relativamente fácil e rápida. O modelo de escolha discreta utilizado para realizar a divisão modal é o modelo logit multinomial. A calibração do modelo logit é realizada com base em 505 observações e considerando três situações: a população de viajantes sem segmentação, com segmentação segundo posse de automóvel no domicílio, e com segmentação segundo a distância de viagem. Analisa-se a precisão da estimativa do número de viagens quando as zonas de controle que formam a área de estudo são divididas segundo o critério de distância de acesso às linhas de ônibus. Na verificação realizada são empregados dados obtidos da pesquisa domiciliar realizada na cidade de Bauru, SP. Os erros cometidos na estimativa são medidos e comparados com os resultados obtidos na pesquisa domiciliar sem expansão. O trabalho demostrou experimentalmente que o procedimento proposto é uma alternativa viável para se obter a distribuição de viagens e, portanto, para determinar a matriz O/D. / A procedure is presented for estimating an origin-destination (O/D) matrix using a discrete choice model jointly with a public transport O/D matrix; the latter being relatively easy and fast to obtain. A multinomial logit discrete choice model is used to determine the mode split of travelers among several available modes. The model is estimated using a sample of 505 observations chosen from a household 0/D survey in the city of Bauru, state of Sao Paulo. This survey is used throughout to verify the results of the proposed methodology. Three different models are considered: one without segmentation of the population, one with segmentation according to car ownership and one with segmentation according to travel distance. An examination is made of the changes on the level of precision of the number of trips estimated with the division of the control zones that conform the study area; the aforementioned division was carried out on the basis of an access distance to the bus network criterion. The estimation errors are measured and compared to the unexpanded O/D survey results. The research shows empirically that the proposed procedure can be used to obtain the distribution of trips and hence to determine an aggregate O/D matrix.

Die vorliegende Arbeit besteht aus vier Aufsätzen, die sich mit der Modellierung von Markenwahlverhalten und Markentreue beschäftigen. Der erste Aufsatz gibt eine Einführung in das Nested Logit Modell und weist auf die Existenz von zwei unterschiedlichen Spezifikationen hin. Das utility maximization nested logit (UMNL) und das non-normalized nested logit (NNNL) Modell besitzen unterschiedliche Eigenschaften, die die Schätzergebnisse beeinflussen. Mit einer Simulationsstudie werden die Konsequenzen der Verwendung verschiedener Softwarepakete demonstriert. Außerdem wird gezeigt, dass nur die UMNL Spezifikation bei Auferlegung einer Parameterrestriktion mit der Zufallsnutzentheorie konform ist. Der zweite Aufsatz untersucht anhand von realen Haushaltspaneldaten den Erfolg einer Familienmarkenstrategie. Die Signaling Theorie liefert einen Rahmen für die dem Markenwahlverhalten zugrunde liegenden psychologischen Prozesse zur Entstehung und Erklärung von produktkategorieübergreifender Markentreue. In einer empirischen Studie wird untersucht, inwieweit in einer Kategorie markentreue Kunden dieser Marke auch in anderen Produktkategorien treu sind. Es wird ein Markentreue-Hebel-Index entwickelt. Im dritten Aufsatz stehen die psychologischen Determinanten von kategorieübergreifenden Zusammenhängen im Markenwahlverhalten im Mittelpunkt. In einer empirischen Studie wird die Risikoaversion als entscheidender Bestimmungsfaktor von kategorieübergreifender Markentreue untersucht. Die konsumentenspezifische Risikoaversion wird dabei über Innovativeness und Status Quo Bias erfasst. Im vierten Aufsatz wird das Hybride Wahlmodell einem breiten Marketingpublikum vorgestellt. Klassische Wahlmodelle gehen davon aus, dass das beobachtbare Verhalten das Resultat eines nicht spezifizierten Evaluationsprozesses des Individuums ist. Der kausalanalytische Ansatz hingegen erlaubt die Spezifikation nicht direkt messbarer Faktoren als latente Variablen und kann somit Wahlmodelle sinnvoll ergänzen. / This thesis is composed of four essays that pick up topics in brand choice and brand loyalty modeling. The first essay gives an introduction to the nested logit model and points attention to the existence of two different specifications. The utility maximization nested logit (UMNL) model and the non-normalized nested logit (NNNL) model have different properties which impact the estimation results. In a simulation study, the consequences of the usage of different software packages for model estimation on the estimation results is demonstrated. It is also shown that only the UMNL specification with an imposed parameter restriction is consistent with the underlying random utility theory. The second essay investigates the success of an umbrella branding strategy using household panel data. Signaling theory provides a framework for the underlying psychological processes in consumers'' brand choice behavior and can contribute in the formation and explanation of loyalty to the brand in multiple categories. An empirical study determines whether there is a tendency for loyal consumers from one product category to be loyal to the same brand in other product categories as well. Therefore, a cross-category brand loyalty leverage index is developed. In the third essay, consumer-specific psychological determinants of cross-category relations between brand loyal choice decisions are discussed. In an empirical study, the concept of risk aversion is considered as the key determinant of cross-category brand loyalty. Consumers'' risk aversion is derived from their innovativeness and status quo bias. In the fourth essay, the hybrid choice model is introduced to the broad marketing audience. Traditional choice models assume that observable behavior results from an unspecified evaluation process of the observed individual. The causal-analytic approach offers the possibility to specify not directly measurable factors as latent variables, and can thus reasonably supplement choice models.

Apresenta-se um procedimento para estimativa da matriz origem-destino (O/D) a partir de um modelo de escolha discreta em combinação com uma matriz O/D de transporte público, a qual pode ser estimada de maneira relativamente fácil e rápida. O modelo de escolha discreta utilizado para realizar a divisão modal é o modelo logit multinomial. A calibração do modelo logit é realizada com base em 505 observações e considerando três situações: a população de viajantes sem segmentação, com segmentação segundo posse de automóvel no domicílio, e com segmentação segundo a distância de viagem. Analisa-se a precisão da estimativa do número de viagens quando as zonas de controle que formam a área de estudo são divididas segundo o critério de distância de acesso às linhas de ônibus. Na verificação realizada são empregados dados obtidos da pesquisa domiciliar realizada na cidade de Bauru, SP. Os erros cometidos na estimativa são medidos e comparados com os resultados obtidos na pesquisa domiciliar sem expansão. O trabalho demostrou experimentalmente que o procedimento proposto é uma alternativa viável para se obter a distribuição de viagens e, portanto, para determinar a matriz O/D. / A procedure is presented for estimating an origin-destination (O/D) matrix using a discrete choice model jointly with a public transport O/D matrix; the latter being relatively easy and fast to obtain. A multinomial logit discrete choice model is used to determine the mode split of travelers among several available modes. The model is estimated using a sample of 505 observations chosen from a household 0/D survey in the city of Bauru, state of Sao Paulo. This survey is used throughout to verify the results of the proposed methodology. Three different models are considered: one without segmentation of the population, one with segmentation according to car ownership and one with segmentation according to travel distance. An examination is made of the changes on the level of precision of the number of trips estimated with the division of the control zones that conform the study area; the aforementioned division was carried out on the basis of an access distance to the bus network criterion. The estimation errors are measured and compared to the unexpanded O/D survey results. The research shows empirically that the proposed procedure can be used to obtain the distribution of trips and hence to determine an aggregate O/D matrix.

This study looked at risk factors associated with vitamin D levels in the body among a representative sample of adults in the U.S., NHANES III (2001-2006) data were used to assess the relationship between several demographic and health risk factors and vitamin D levels in the body. The Baseline-Category Logit Model was used to test the association between vitamin D level and the potential risk factors age, education, ethnicity, poverty status, physical activity, smoking, alcohol, obesity, diabetes and total cholesterol with both genders. Vitamin D insufficiency and deficiency were significantly associated with age, race, education, physical activity, obesity, diabetes and total cholesterol level for both genders. Almost half of the adults sampled in these data had vitamin D levels lower than the recommended limits, with the highest frequency among the younger groups. Determining an individual’s vitamin D level is very difficult without proper clinical testing. Many of those who have low vitamin D levels are unaware. With such a high prevalence of individuals with low vitamin D levels in the U.S. and a better understanding of characteristics associated with these lower levels, increased education and prevention efforts should be focused toward those with higher risk characteristics.

由於傳統國人的觀念「安土重遷、購屋保值」，使得一般人都想要擁有自己的房屋，然而房價居高不下，國人大多需仰賴長時間之儲蓄，始有購屋之能力，使得許多人一屋難求，因此，本研究利用79年台北市家戶及住宅普查調查資料，探討家戶住宅租購選擇之影響因素，並瞭解選擇租屋或購屋之對象及其選擇行為，進而估計各類型家戶選擇租屋或購屋之機率。 本研究採用巢式多項羅吉特(Logit)模型分析家戶之住宅租購選擇，及各類型家戶之租購選擇機率，並估計出自有房屋市場與租賃房屋市場間有很高的替代性存在，若忽略租購選擇，則所估計到之住宅需求將會造成誤差，因此，為避免產生錯誤的結果，不應將自有房屋市場與租賃房屋市場分開來討論。 在影響家戶住宅租購選擇之因素方面，所得、住宅價格．利率、家庭因素、預期因素、心理因素等其他因素，皆會影響家戶之選擇行為，而本研究僅就家戶組成來探討住宅之租購選擇。此外，在住宅租購選擇之機率方面，各類型家戶選擇自有房屋之機率均相當高，凸顯台北市高住宅自有率之現象，此種現象正反映傳統國人購屋保值的觀念，房價在高漲的情況下，許多人仍將購屋視為理想目標。

傳統的犯罪學在於研究犯罪者生理與心理的特性以及外在環境因素，藉以 尋求預防犯罪的策略。其實，預防犯罪的目的是要減少被害者數目。因此 ，吾人希望藉由探討被害者家庭形態的特質與社會犯罪的關係，使防範被 害的政策能夠落實到家庭層面。所以本研究的目的就是要利用統計方法， 透過數字及圖形上的表達，來描述住戶之家計負責人的特性(如性別、年 齡、教育程度)與住戶遭受侵害的事實(曾或否)及類型(如財物侵害、暴力 侵害、其他侵害)之間的關係，期為犯罪統計工作提供一些分析的工具。 在本篇論文，首先利用對數線性模式分析，選出最適合的模式，以解釋家 計負責人之性別、年齡、教育程度與住戶是否遭受侵害等四個變數之間的 關聯性。接著利用推理值模式分析探討家計負責人之性別，年齡，教育程 度這三個解釋變數各分級對住戶遭受侵害之影響的主效應及交互作用效應 。最後則是利用對應分析處理資料，透過對應分析的顯示圖形，可以用來 解釋家計負責人之性別、年齡、教育程度以及住戶遭受侵害的犯罪類型之 間的對應關係。

2007年8月美國爆發次貸危機(Subprime Crisis)，如此新型態的金融危機是否可由金融危機預警系統預測？是本文所欲探討的目標。本文採用訊號方法、固定效果下的Panel Logit Model和CART(Classification and Regression Tree)三種計量方法建構危機預警模型。最後利用美國2006年至2008年資料，驗證本文所建構之預警模型是否能夠有效預測次貸危機的發生。 / “Could banking early warning systems help to predict Sub-prime crisis?” That is the main issue that we want to discuss. We combine three kinds of early warning systems models – Signal Approach, fixed effect panel logit model, and CART approach – to create a new banking early warning system(EWS). We will use the US 2006-2008 data to examine whether this new EWS could predict the Sub-prime crisis correctly.

股利政策之選擇攸關公司資金成本效益與投資者權益，投資者依據公司未來股利發生時點、金額與不確定性等因素來判斷公司的內涵價值（intrinsic value），亦即股利的配發是公司管理當局與股東利益間的抵換。本論文將考量我國企業實務上股利支付內容，將股利形式細分為無股利分配及有股利配發，包含單一股利政策－現金股利、盈餘配股、公積配股，複合股利政策－現金股利搭配盈餘配股、現金股利搭配公積配股、盈餘配股搭配公積配股及綜合股利政策（現金股利、盈餘配股、公積配股三者兼發）之發放模式，共計有八種股利支付類型。 　　共同基金乃證券市場中連接有價證券與投資者間之專業投資管道，由於基金收益直接進行再投資產生之租稅規避濾網（tax-free filter）效果，藉以探討公司選擇股利政策時之顧客效應。同時將影響股利政策制定之構面因素－公司的股利配發來源、獲利性狀況、資金融通與債務償還能力、企業未來投資成長機會、產業類別與租稅改革制度等一併納入分析檢證。 　　本論文採用實證模型分析為主，文獻檢閱分析為輔，以1995- 2000年我國上市公司曾被國內股票型投信基金選為投資標的者，運用Mutinomial logit model及股利政策相關文獻，建構本研究實證模型；在政策意涵上，股利稅制的規劃與共同基金租稅課徵問題亦一併討論，以檢討我國現行稅制環境是否健全。 　　本研究所獲結論歸納如下： 　　(1) 共同基金持股比重變動對股利政策影響存在明顯差異 　　　　股票型基金持股比重大小對現金股利搭配盈餘配股與盈餘配股搭配公積配股選擇皆有顯著正效果，但是否基於股東特性而論則難以斷定。現行我國共同基金持股率占公司股權結構比重不大、透過除權日規避股利發放及大股東利用共同基金進行投信交叉持股等問題，皆是造成實證結果與預期相左之可能原因。 　　(2) 租稅制度改革對股利配發內容的選擇扮演要角 　　　　兩稅合一制度實施後，總股利支付率較實施前為高，主要在於規避保留盈餘加徵10﹪的稅負及股利的發放具有抵稅權，使公司不再傾向保留盈餘而提高發放股利之意願。 　　(3) 股利配發形式選擇 　　　　◎支持股利配發來源與股利政策的關係 　　　　◎公司舉債程度高低與現金股利配發機率呈負相關 　　　　◎公司獲利性狀況攸關股利發放與否 　　　　◎未來成長機會與現金股利配發呈反向關係 　　　　◎股利政策具有產業差異性，高科技產業傾向選擇股票股利偏 高之股利組合政策 　　以公司立場而言，宜需定期評估經營利基與股利政策選擇因素，建立一套合理之股利預測分析模式，並考量平衡股利政策（現金股利與股票股利搭配）模式，以減緩我國企業普遍採股票股利配發造成股權過度膨脹及盈餘稀釋之弊。從投資者角度來看，共同基金係專業化投資管道，藉由大眾資金匯集之規模效益，能使資金有效配置（或風險分散）並提升國內專業法人比重，減緩市場投機風氣。然共同基金投資相關法令之設立仍需更臻完備，稅制環境對公司股利決策具重要影響性，對追求租稅中立目標而言，證券交易所得是否復徵所得稅，亦是政府單位需審慎考量的議題。

Southeastern states have about 26 percent of the nations total fatalities, and are about 24 percent above the national mean over recent years. Descriptive statistics, graphs, and figures are used to illustrate and quantify the crash trends, which depict a comprehensive picture of status and trends of the fatal crashes in southeastern states. The severity of crashes is studied as a function of characteristics of the person involved in the crash, vehicle, traffic condition, physical road geometry, and environmental factors. Detailed geometric feature data were collected for this study, which makes it possible to investigate the relationship between geometric features and crash severity. This study identifies causal factors contributing to the high fatality rate in southeastern states, and sheds light on the differences and similarities among these states for reducing the severity of fatal crashes, by developing multinomial logit models to explain the severity and type of fatal crashes.

In both developed and emerging-economies, major cities continue to experience increasing traffic congestion. To address this issue, complex Traffic Management Systems (TMS) are employed in recent years to help manage traffic. These systems fuse traffic-surveillance-related information from a variety of sensors deployed across traffic networks. A TMS requires real-time information to make effective control decisions and to deliver trustworthy information to users, such as travel time, congestion level, etc. There are three fundamental inputs required by TMS, namely, traffic volume, vehicular speed, and traffic density. Using conventional traffic loop detectors one can directly measure flow and velocity. However, traffic density is more difficult to measure. The situation becomes more difficult for multi-lane motorways due to drivers lane-change behaviour. This research investigates statistical modelling and analysis of traffic flow. It contributes to the literature of transportation and traffic management and research in several aspects. First, it takes into account lane-changes in traffic modelling through incorporating a Markov chain model to describe the drivers lane-change behaviour. Secondly, the lane change probabilities between two adjacent lanes are not assumed to be fixed but rather they depend on the current traffic condition. A discrete choice model is used to capture drivers lane choice behaviour. The drivers choice probabilities are modelled by several traffic-condition related attributes such as vehicle time headway, traffic density and speed. This results in a highly nonlinear state equation for traffic density. To address the issue of high nonlinearity of the state space model, the EKF and UKF is used to estimate the traffic density recursively. In addition, a new transformation approach has been proposed to transform the observation equation from a nonlinear form to a linear one so that the potential approximation in the EKF & UKF can be avoided. Numerical studies have been conducted to investigate the performance of the developed method. The proposed method outperformed the existing methods for traffic density estimation in simulation studies. Furthermore, it is shown that the computational cost for updating the estimate of traffic densities for a multi-lane motorway is kept at a minimum so that online applications are feasible in practice. Consequently the traffic densities can be monitored and the relevant information can be fed into the traffic management system of interest.