Interorganizational alliances are widely recognized as critical to product innovation. A notable trend is the rapid growth of new product development (NPD) alliances between large, well-established firms and small, growing firms. This dissertation is comprised of two studies on the formation and termination of asymmetric new product development alliances. In study one I examine the factors that drive the changes in shareholder values of the partner firms. I develop and empirically test a model of short-term changes in shareholder values of larger and smaller firms involved in NPD alliances, using the event study methodology on data covering 167 asymmetric alliances in the information technology and communication industries. The model accounts for selection correction, potential cross-correlation across the residuals from the models of firm value changes for the larger and smaller firms, and unobserved heterogeneity. The results suggest that both the partners experience significant short-term financial gains, but there are considerable asymmetries between the larger and smaller firms with regard to the effects of alliance, partner and firm characteristics on the gains of the partner firms. The findings of this study have important implications for managers of both large and small firms. In study two I develop and test a framework of the determinants of new product alliance (NPA) terminations. The hypotheses for study two are tested on a unique database comprised of 401 new product alliances involving 24 pharmaceutical firms during 1990-2005. NPA terminations are modeled using Cox’s proportional hazard specification that accounts for the unobserved heterogeneity of firms with multiple NPAs, competing risks and ties among NPA duration times. The results suggest that NPA terminations are not made in isolation but are influenced by composition of the firm’s portfolio. The results also suggest that NPA terminations are predicted to a great extent by competition between alliances (i.e., product market rivalry) and competition within alliances (i.e., partner value). The findings of this study have important implications for managing a portfolio of new product partnerships.

Interorganizational alliances are widely recognized as critical to product innovation. A notable trend is the rapid growth of new product development (NPD) alliances between large, well-established firms and small, growing firms. This dissertation is comprised of two studies on the formation and termination of asymmetric new product development alliances. In study one I examine the factors that drive the changes in shareholder values of the partner firms. I develop and empirically test a model of short-term changes in shareholder values of larger and smaller firms involved in NPD alliances, using the event study methodology on data covering 167 asymmetric alliances in the information technology and communication industries. The model accounts for selection correction, potential cross-correlation across the residuals from the models of firm value changes for the larger and smaller firms, and unobserved heterogeneity. The results suggest that both the partners experience significant short-term financial gains, but there are considerable asymmetries between the larger and smaller firms with regard to the effects of alliance, partner and firm characteristics on the gains of the partner firms. The findings of this study have important implications for managers of both large and small firms. In study two I develop and test a framework of the determinants of new product alliance (NPA) terminations. The hypotheses for study two are tested on a unique database comprised of 401 new product alliances involving 24 pharmaceutical firms during 1990-2005. NPA terminations are modeled using Cox’s proportional hazard specification that accounts for the unobserved heterogeneity of firms with multiple NPAs, competing risks and ties among NPA duration times. The results suggest that NPA terminations are not made in isolation but are influenced by composition of the firm’s portfolio. The results also suggest that NPA terminations are predicted to a great extent by competition between alliances (i.e., product market rivalry) and competition within alliances (i.e., partner value). The findings of this study have important implications for managing a portfolio of new product partnerships.

Η εξέλιξη των επιχειρήσεων και η ανάπτυξή τους έχει απασχολήσει από παλιά την ακαδημαϊκή κοινότητα (Gibrat 1931). Ένα εξιδεικευμένο τμήμα της βιβλιογραφίας αυτής ασχολείται με τον προσδιορισμό των παραγόντων επιβίωσης των επιχειρήσεων, που αποτελεί αντικείμενο της παρούσας διατριβής. Οι παράγοντες αυτοί μπορεί να αναφέρονται σε ίδια χαρακτηριστικά των επιχειρήσεων και του κλάδου στον οποίο ανήκουν ή και σε μακροοικονομικούς παράγοντες. Ο προαναφερθείς σκοπός επιτυγχάνεται μέσα από την ανάπτυξη των θεωρητικών υποθέσεων με βάση τις οποίες καταρτίζεται το οικονομετρικό υπόδειγμα. Το κύριο βάρος της εργασίας εστιάζεται στην ανάλυση των συνεχών υποδειγμάτων κινδύνου και το έλεγχο των υποθέσεων στα οποία βασίζονται. Ειδικότερα εξετάζεται η επίδραση της μορφής των δεδομένων και της χρονικής συνάθροισης στους εκτιμητές των εφαρμοζόμενων στην μέχρι σήμερα εμπειρική έρευνα υποδειγμάτων. Συγκρίνοντας διαφόρους τύπους διακριτών και συνεχών υποδειγμάτων κινδύνου οδηγούμεθα στο συμπέρασμα ότι τα μέχρι σήμερα εφαρμοζόμενα συνεχή υποδείγματα καταλήγουν σε μεροληπτικά αποτελέσματα. Η εφαρμογή διακριτών υποδειγμάτων κινδύνου PCE, που αποτελεί και την συνεισφορά μας στη βιβλιογραφία, περιορίζει σε σημαντικό βαθμό τα μειονεκτήματα των εφαρμοζόμενων στην εμπειρική έρευνα υποδειγμάτων. Η εφαρμογή της παραπάνω μεθοδολογίας προϋποθέτει δεδομένα τύπου longitudinal τα οποία δεν είναι συνήθως διαθέσιμα. Αυτό εξηγεί το γεγονός των περιορισμένων εμπειρικών εργασιών στο χώρο της επιβίωσης των επιχειρήσεων. Δεδομένης της δυσκολίας εξεύρεσης κατάλληλων δεδομένων η εμπειρική εφαρμογή στηρίχθηκε σε δεδομένα δημοσίων εγγραφών της ελληνικής κεφαλαιαγοράς για την περίοδο 1993-2002. Από όσο είμαστε σε θέση να γνωρίζουμε είναι η πρώτη φορά στη βιβλιογραφία που διεξάγεται παρόμοια έρευνα με ελληνικά δεδομένα δημοσίων εγγραφών. Τέλος, η διατριβή καταλήγει στην εξαγωγή των βασικών συμπερασμάτων και προτάσεων σχετικά με μελλοντική έρευνα. / Most economic phenomena are measured over long time intervals and this naturally leads us to see time as a discrete variable. More precisely, the underlying duration process occurs in continuous time, while the observed data comes in grouped form. In applied research, while firm life data are grouped into time intervals, continuous hazard models are used. The purpose of this paper is to investigate the sensitivity of regression coefficients and the coefficient duration dependence of the discrepancy between the statistical model and the data generating process.

Detailed mapping and granulometric analyses of the 1992 Cerro Negro tephra blanket reveal remarkable departures from the expected distribution of tephra. Isomass maps show that the major axis of dispersion for the eruption was to the SW of the cone and that the coarser-grained particles, ranging from -4.0 -- 1.0 f, were deposited primarily along the major axis of dispersion with deposits thinning off of the axis. Comparable isomass maps for finer-grained particles, 1.5 - 3.5 f, show that these particles were primarily deposited along the edges of the deposit, off of the major axis of dispersion. Advection-diffusion models for tephra fallout currently widely used in volcanology do not account for this deposition pattern. Rather, it appears that interaction between the wind field, which developed a strong cross flow during the eruption, and the ascending tephra plume resulted in the formation of turbulent structure in the plume. Particles with a settling velocity greater than ~1-2m/s (diameter >0.5 mm) were able to overcome the turbulent structure and settled in a manner predicted by the advection-diffusion equation. Those with lower settling velocities were caught up in turbulent structure and deposited off of the major axis of dispersion, near the edges of the overall tephra blanket. Thus, this data set provides the first estimate of the strength of such turbulent structures in advecting plumes, and illustrates the limitations of the typical advection-diffusion models in describing some transport processes.

This paper examines the relationship between the accounting conservatism construct and the prediction of corporate bankruptcy. Prior research has explored the link between accounting quality and bankruptcy prediction, but it has not examined the relationship between accounting conservatism and bankruptcy prediction. This study hypothesizes that the inclusion of conservatism metrics in the bankruptcy hazard model estimation process should have an incremental effect on the predictive ability of bankruptcy hazard models. This paper finds that the inclusion of conservatism metrics does enhance the predictive power of bankruptcy hazard models for certain subgroups of a population partitioned on the basis of accounting conservatism metrics.

Acute respiratory infections (ARI) caused by influenza and other respiratory viruses affect millions of people annually. Although usually self-limiting a more complicated or severe course may occur in previously healthy people but are more likely in individuals with underlying illnesses. The most common viral agent is rhinovirus whereas influenza is less frequent but is well known to cause winter epidemics. In primary care, rapid diagnosis of influenza virus infections is essential in order to provide treatment. Clinical presentations vary among the different pathogens but may overlap and may also depend on host factors. Predictive models have been developed for influenza but study results may be biased because only individuals presenting with fever were included. Most of these models have not been adequately validated and their predictive power, therefore, is likely overestimated. The main objective of this thesis was to compare different mathematical models for the derivation of clinical prediction rules in individuals presenting with symptoms of ARI to better distinguish between influenza, influenza A subtypes and entero-/rhinovirus-related illness in children and adults and to evaluate model performance by using data-splitting for internal validation. Data from a completed prospective cluster-randomized trial for the indirect effect of influenza vaccination in children of Hutterite communities served as a basis of my thesis. There were a total of 3288 first episodes per season of ARI in 2202 individuals and 321 (9.8%) influenza positive events over three influenza seasons (2008-2011). The data set was divided into children under 18 years and adults. Both data sets were randomly split by subjects into a derivation (2/3 of the dataset) and a validation population (1/3 of the dataset). All predictive models were developed in the derivation sets. Demographic factors and the classical symptoms of ARI were evaluated with logistic regression and Cox proportional hazard models using forward stepwise selection applying robust estimators to account for non-independent data and by means of recursive partitioning. The beta coefficients of the independent predictors were used to develop different point scores. These scores were then tested in the validation groups and performance between validation and derivation set was compared using receiver operating characteristics (ROC) curves. We determined sensitivities and specificities, positive and negative predictive values, and likelihood ratios at different cut-points which could reflect test and treatment thresholds. Fever, chills, and cough were the most important predictors in children whereas chills and cough but not fever were most predictive of influenza virus infection in adults. Performance of the individual models was moderate with areas under the receiver operating characteristic curves between 0.75 and 0.80 for the main outcome influenza A or B virus infection. There was no statistically significant difference in performance between the derivation and validation sets for the main outcome. The results have shown, that various mathematical models have similar discriminative ability to distinguish influenza from other respiratory viruses. The scores could assist clinicians in their decision-making. However, performance of the models was slightly overestimated due to potential clustering of data and the results would first needed to be validated in a different population before application in clinical practice. / Thesis / Master of Science (MSc) / Every year, millions of people are attacked by "the flu" or the common cold. Certain signs and symptoms apparently are more discriminative between the common cold and the flu. However, the decision between starting a simple symptom orientated treatment, treating empirically for influenza or ordering a rapid diagnostic test that has only moderate sensitivity and specificity can be challenging. This thesis, therefore, aims to help physicians in their decision-making process by developing simple scores and decision trees for the diagnosis of influenza versus non-influenza respiratory infections. Data from a completed trial for the indirect effect of influenza vaccination in children of Hutterite communities served as a basis of my thesis. There were a total of 3288 first seasonal episodes of ARI in 2202 individuals and 321 (9.8%) influenza positive events over three influenza seasons (2008-2011). The data set was divided into children under 18 years and adults. Both data sets were split into a derivation and a validation set (=holdout group). Different mathematical models were applied to the derivation set and demographic factors as well as the classical symptoms of ARI were evaluated. The scores generated from the most important factors that remained in the model were then tested in the validation group and performance between validation and derivation set was compared. Accuracy was determined at different cut-points which could reflect test and treatment thresholds. Fever, chills, and cough were the most important predictors in children whereas chills and cough but not fever were most predictive of influenza virus infection in adults. Performance of the individual models was moderate for the main outcome influenza A or B virus infection. There was no statistically significant difference in performance between the derivation and validation sets for the main outcome. The results have shown, that various mathematical models have similar discriminative ability to distinguish influenza from other respiratory viruses. The scores could assist clinicians in their decision-making. However, the results would first needed to be validated in a different population before application in clinical practice.

In direct marketing, understanding the response behavior of consumers to marketing initiatives is a pre-requisite for marketers before implementing targeting strategies to reach potential as well as existing consumers in the future. Consumer response can either be in terms of the incidence or timing of purchases, category/ brand choice of purchases made as well as the volume or purchase amounts in each category. Direct marketers seek to explore how past consumer response behavior as well as their targeting actions affects current response patterns. However, considerable heterogeneity is also prevalent in consumer responses and the possible sources of this heterogeneity need to be investigated. With the knowledge of consumer response and the corresponding heterogeneity, direct marketers can devise targeting strategies to attract potential new consumers as well as retain existing consumers. In the first essay of my dissertation (Chapter 2), I model the response behavior of donors in non-profit charity fund-raising in terms of their timing and volume of donations. I show that past donations (both the incidence and volume) and solicitation for alternative causes by non-profits matter in donor responses and the heterogeneity in donation behavior can be explained in terms of individual and community level donor characteristics. I also provide a heuristic approach to target new donors by using a classification scheme for donors in terms of the frequency and amount of donations and then characterize each donor portfolio with corresponding donor characteristics. In the second essay (Chapter 3), I propose a more structural approach in the targeting of customers by direct marketers in the context of customized retail couponing. First I model customer purchase in a retail setting where brand choice decisions in a product category depend on pricing, in-store promotions, coupon targeting as well as the face values of those coupons. Then using a utility function specification for the retailer which implements a trade-off between net revenue (revenue – coupon face value) and information gain, I propose a Bayesian decision theoretic approach to determine optimal customized coupon face values. The optimization algorithm is sequential where past as well as future customer responses affect targeted coupon face values and the direct marketer tries to determine the trade-off through natural experimentation. / text

Assumir suposições especiais sobre a função de risco tem sido a estratégia adotada por vários autores, com intuito de garantir modelos gerais e abrangentes, tanto para a análise de dados de sobrevivência quanto de conDabilidade. Neste estudo, modelos aplicados a dados da área de sobrevivência e conDabilidade são considerados. A Dnalidade deste estudo é propor modelos mais Pexíveis e/ou mais abrangentes de forma a generalizar modelos já existentes, bem como estudar suas propriedades e propor possíveis comparações entre os modelos via testes de hipóteses. Considera-se nesta tese, três classes de modelos baseados na função de risco (modelos de risco). A primeira classe apresenta-se como um caso particular do modelo de risco estendido (Louzada-Neto, 1999), formada por modelos que relacionam o parâmetro de escala a covariáveis, sendo que esse relacionamento pode ser considerado log-linear ou log-nãolinear. Considera-se um modelo particular onde a dependência do parâmetro de escala se dá de forma log-não-linear. Na segunda classe considera-se modelos que estão vinculados a dados de riscos competitivos, quando se tem ou não informação sobre qual tipo de risco foi responsável pela falha de um equipamento ou pelo óbito de um paciente. A terceira classe de modelos foi proposta, nesta tese, relacionando o contexto de modelos de longa duração. / Assuming special suppositions for the hazard function have been the strategy used for many authors in order to guarantee general and Pexible models for survival and reliability data. The present thesis considers two classes of hazard models, with the basic objective of proposing more Pexible models, studying their properties and proposing possible comparisons via hypothesis tests. We consider, three families of models where the struture was based in hazard function. The Drst class is a special case of the extented hazard model (Louzada, 1999). This class of models is composed by models with relationship between the scale parameter and the covariates could be log-linear or log-non-linear, we consider the log-non-linear. The second class is into the context of competing risk, where we do not known what kind of risk is responsable for the fail.or death. The third class, proposed in this work refers to a context of long term survivals. All the procedures were ilustrated in real datasets

Assumir suposições especiais sobre a função de risco tem sido a estratégia adotada por vários autores, com intuito de garantir modelos gerais e abrangentes, tanto para a análise de dados de sobrevivência quanto de conDabilidade. Neste estudo, modelos aplicados a dados da área de sobrevivência e conDabilidade são considerados. A Dnalidade deste estudo é propor modelos mais Pexíveis e/ou mais abrangentes de forma a generalizar modelos já existentes, bem como estudar suas propriedades e propor possíveis comparações entre os modelos via testes de hipóteses. Considera-se nesta tese, três classes de modelos baseados na função de risco (modelos de risco). A primeira classe apresenta-se como um caso particular do modelo de risco estendido (Louzada-Neto, 1999), formada por modelos que relacionam o parâmetro de escala a covariáveis, sendo que esse relacionamento pode ser considerado log-linear ou log-nãolinear. Considera-se um modelo particular onde a dependência do parâmetro de escala se dá de forma log-não-linear. Na segunda classe considera-se modelos que estão vinculados a dados de riscos competitivos, quando se tem ou não informação sobre qual tipo de risco foi responsável pela falha de um equipamento ou pelo óbito de um paciente. A terceira classe de modelos foi proposta, nesta tese, relacionando o contexto de modelos de longa duração. / Assuming special suppositions for the hazard function have been the strategy used for many authors in order to guarantee general and Pexible models for survival and reliability data. The present thesis considers two classes of hazard models, with the basic objective of proposing more Pexible models, studying their properties and proposing possible comparisons via hypothesis tests. We consider, three families of models where the struture was based in hazard function. The Drst class is a special case of the extented hazard model (Louzada, 1999). This class of models is composed by models with relationship between the scale parameter and the covariates could be log-linear or log-non-linear, we consider the log-non-linear. The second class is into the context of competing risk, where we do not known what kind of risk is responsable for the fail.or death. The third class, proposed in this work refers to a context of long term survivals. All the procedures were ilustrated in real datasets

Esta tesis contribuye a la literatura analizando los efectos causales que el estado de salud tiene sobre la participación laboral en la población en edad de trabajar. De este modo, analiza los efectos que un deterioro en el estado de salud tiene sobre la participación laboral del individuo, así como los efectos de proveer cuidados informales sobre la participación laboral femenina. El primer capítulo utiliza una aproximación empírica homogénea en nueve países europeos, lo que hace posible relacionar las diferencias encontradas con diferencias en el contexto institucional. El segundo capítulo analiza el papel que juega el estado de salud en las transiciones hacia y fuera del empleo. Los resultados muestran que el estado de salud general afecta simétricamente las entradas y salidas del empleo, mientras que cambios en el estado de salud mental sólo influyen el riesgo de abandonar el empleo. El tercer capítulo examina los efectos de varios tipos de cuidados informales en el comportamiento laboral femenino. Los resultados sugieren que los costes de oportunidad laborales aparecen en aquellas mujeres que conviven con la persona dependiente, al mismo tiempo que los efectos negativos surgen cuando se proveen cuidados informales por un período superior al año. / This thesis aims to contribute to the literature with an attempt to identify the causal effects of health on labour market outcomes in the working-age population. I analyse the effects of the onset of a health shock on the individuals' labour market outcomes, and also the effects of caregiving on female labour participation. The first chapter uses a homogeneous empirical framework to estimate the first set of effects on nine European countries, which allows me to relate the empirical estimates to differences in social security arrangements across these countries. The second chapter analyses the role of health in exits out of and entries into employment and the results show that general health affects symmetrically entries into and exits out of employment, but changes in mental health status influence only the hazard of non-employment for the stock sample of workers. The third chapter examines the effects of various types of informal care on female labour behaviour and the results suggest the existence of labour opportunity costs for those women who live with the dependent person they care for, and the negative effects appear when caregiving for more than a year.