This thesis aimed to apply the causal framework with potential outcomes to examine the causal effect of air temperature on reported violent crimes in Swedish municipalities. The Generalized Estimating Equations method was used on yearly, monthly and also July only data for the time period 2002-2014. One significant causal effect was established but the majority of the results pointed to there being no causal effect between air temperature and reported violent crimes.

Ingeniera Civil Industrial / La industria del retail en Chile en los últimos años ha enfrentado una disminución en las altas tasas de crecimiento que había acostumbrado a tener, haciéndose evidente el desafió de adaptarse a un mercado turbulento, que lleva consigo un cambio constante de las expectativas de los consumidores, cuya preferencia puede significar el éxito o el fracaso de cualquier empresa. Y así lo entienden las empresas de tiendas por departamento, donde su preocupación ha ido en aumento respecto a la obtención y retención de sus clientes. El objetivo de este trabajo es el de diseñar y proponer una estrategia de email marketing que permita aumentar la probabilidad de pertenecer a la categoría de alto valor de los clientes, mediante la aplicación de acciones promocionales que sean capaces de aumentar los niveles de visitas y/o los niveles de venta de los clientes. Para lograr esto, se comenzará por analizar aquellas variables que podrían ser relevante para explicar la probabilidad de cada cliente de pertenecer a esta categoría. Con estas variables se procederá a diseñar un modelo que sea capaz de calcular dicha probabilidad, conocida como Propensity Score. Con esta información se realizará un estudio experimental que busca probar distintos tipos de promociones (distintos montos y porcentajes de descuento) y medir su efectividad en términos de contribución de visitas como de ventas. Para alcanzar los objetivos mencionados, se trabajará con clientes identificables, es decir, clientes con Rut, con registro de su información demográfica y de su comportamiento de compra. A estos se les aplicaran herramientas de estadística descriptiva, técnicas RFM, comparación de modelos de propensión mensual y la aplicación de un diseño experimental. El principal entregable de esta memoria corresponde a una propuesta de gestión de clientes de alto valor, clientes F, la cual señala qué promociones son más adecuadas para aumentar la propensión de cada cliente, sus niveles de ventas y/o sus niveles de visitas. La propuesta diferenciará entre distintos segmentos de clientes según su tipo (cliente F o normal), según su nivel de venta y según su nivel de visitas. Las principales conclusiones del estudio son que promociones enviadas por canal POS no son capaces de aumentar la propensión del cliente. Sin embargo, promociones como el 40% de dcto. sí aumentaron la probabilidad de los clientes sin importar el segmento al cual pertenecían, y mejor aún, mostraron ser rentables para la empresa. Se propone como estudios futuros, probar otros canales adicionales al cupón POS y email, los que quizás tengan mayor efectividad al momento de hacer que un cliente se sienta importante y decida finalmente acercarse a la tienda a comprar. / 16/03/2022

Background: Although palliative chemotherapy is the standard of care for patients diagnosed with stage IV NSCLC, patients with a small metastatic burden, “oligometastatic” disease, may benefit from definitive local therapy. Methods: We identified 186 patients (26% of Stage IV patients) prospectively enrolled in our institutional database from 2002-2012 with oligometastatic disease, which we defined as five or fewer distant metastatic lesions at diagnosis. Univariable and multivariable Cox proportional hazards models were used to identify patient and disease factors associated with improved survival. Using propensity score methods, we investigated the effect of definitive local therapy to the primary site on overall survival. Results: Median age at diagnosis was 61 years, 51% of patients were female, 12% had squamous histology, and 33% had N0-1 disease. On multivariable analysis, ECOG performance status ≥2 (hazard ratio [HR] 2.43), nodal status N2-3 (HR 2.16), squamous pathology, and metastases to multiple organs (HR 2.11) were associated with a greater hazard of death (all p<0.01). Number of metastatic lesions and size of primary were not significantly associated with overall survival. Definitive local therapy to the primary tumor was associated with prolonged survival (HR 0.65, p=0.043). Conclusions: Definitive local therapy to the primary tumor appears to be associated with improved survival in patients with oligometastatic NSCLC. Select patient and tumor characteristics, including good performance status, non-squamous histology, and limited nodal disease, may predict for improved survival in these patients.

Previous research implementing stratification on the propensity score has generally relied on using five strata, based on prior theoretical groundwork and minimal empirical evidence as to the suitability of quintiles to adequately reduce bias in all cases and across all sample sizes. This study investigates bias reduction across varying number of strata and sample sizes via a large-scale simulation to determine the adequacy of quintiles for bias reduction under all conditions. Sample sizes ranged from 100 to 50,000 and strata from 3 to 20. Both the percentage of bias reduction and the standardized selection bias were examined. The results show that while the particular covariates in the simulation met certain criteria with five strata that greater bias reduction could be achieved by increasing the number of strata, especially with larger sample sizes. Simulation code written in R is included.

京都大学 / 0048 / 新制・課程博士 / 博士(医学) / 甲第22311号 / 医博第4552号 / 新制||医||1040(附属図書館) / 京都大学大学院医学研究科医学専攻 / (主査)教授 福原 俊一, 教授 坂井 義治, 教授 妹尾 浩 / 学位規則第4条第1項該当 / Doctor of Medical Science / Kyoto University / DFAM

Adjusting for baseline pre-intervention characteristics between treatment groups, through the use of propensity score matching methods, is an important step that enables researchers to do causal inference with confidence. This is critical, largely, due to the fact that practical treatment allocation scenarios are non-randomized in nature, with various inherent biases that are inevitable, and therefore requiring such experimental manipulations. These propensity score matching methods are the available tools to be used as control mechanisms, for such intrinsic system biases in causal studies, without the benefits of randomization (Lane, To, Kyna , & Robin, 2012). Certain assumptions need to be verifiable or met, before one may embark on a propensity score matching causal effects journey, using the Rubin causal model (Holland, 1986), of which the main ones are conditional independence (unconfoundedness) and common support (positivity). In particular, with this dissertation we are concerned with elaborating the applications of these matching methods, for a ‘strong-ignorability’ case (Rosenbaum & Rubin, 1983), i.e. when both the overlap and unconfoundedness properties are valid. We will take a journey from explaining different experimental designs and how the treatment effect is estimated, closing with a practical example based on two cohorts of enrolled introductory statistics students prior and post-clickers intervention, at a public South African university, and the relevant causal conclusions thereof. Keywords: treatment, conditional independence, propensity score, counterfactual, confounder, common support / Dissertation (MSc)--University of Pretoria, 2020. / Statistics / MSc / Unrestricted

Policies delegating control of educational policy to the local level are widespread, yet there has been little examination of the effects of such distributed decision making in the area of advanced education programming. We used propensity score matching to examine the effectiveness of locally developed policies for identifying intellectually gifted children identifying themselves as Black or from low-socioeconomic backgrounds across one large U.S. state (Florida) that has a state-level gifted education mandate. Ongoing underrepresentation of traditionally marginalized groups in gifted education was evident, even among districts with policies specifically designed to ameliorate disproportional representation. However, the presence of such a policy reduced the degree of underrepresentation.

Objective: The study assesses the impact of dual enrollment participation on remediation and completion for traditional first time, full-time freshmen at a community college in Northeast Tennessee. Method: This study began with the full population of 1,232 students who enrolled between 2008 and 2012 at a community college in northeast Tennessee the fall semester after finishing high school. The population was required to have American College Testing (ACT) scores, completely fill out the Free Application for Federal Student Aid (FAFSA), enroll full-time as a degree-seeking student, and complete the first fall semester. Propensity score matching was utilized to eliminate self-selection bias and enable parametric comparisons using optimal matching of dual enrollment participants and non-participants while controlling for a range of covariates. Results: The analyses showed that community college students who participated in dual enrollment were (a) 9% or nearly 3.4 times less likely to take remediation, (b) 26% or nearly 2.5 times more likely to graduate in 2 years, and (c) 28% or nearly 1.5 times more likely to graduate in 3 years. Contributions: This study contributes to the literature showing that dual enrollment reduces remediation rates and assists in timely completions for community college students. Policy recommendations are to increase equitable participation, normalize dual enrollment for students academically able to do college coursework, align state terminology with the nation, and improve data for future research.

京都大学 / 0048 / 新制・課程博士 / 博士(社会健康医学) / 甲第19636号 / 社医博第69号 / 新制||社医||9(附属図書館) / 32672 / 京都大学大学院医学研究科社会健康医学系専攻 / (主査)教授 佐藤 俊哉, 教授 今中 雄一, 教授 小西 郁生 / 学位規則第4条第1項該当 / Doctor of Public Health / Kyoto University / DFAM

No description available.