The process of research is fraught with rote terminology that, when used blindly, can bend our methodological actions away from our theoretical intentions. This investigation is aimed at developing two methods for bringing meaning and interpretability to research when we work with confounds. I argue, with the first method, that granting confounds substantive influence in a network of related variables (rather than viewing confounds as nuisance variables) enhances the conceptual dimension with which phenomena can be explained. I evaluated models differing in how confounds were specified using data from the Survey of Health, Ageing and Retirement in Europe (SHARE). Generally, minor alterations to model specifications, such as direction of causal pathways, did not change model parameter estimates; however, the conceptual meaning of how the confounds interacted with other variables in the model changed drastically. Another frequent misconceptualization of confounds, detailed by the second method, occurs when confounds are used as proxy variables to control for variance that is not directly measureable, and no explicit attempt is made to ensure that the proxy variable adequately represents the underlying, intended construct. For this second demonstration, I used SHARE data to estimate models varying in the degree to which proxy variables represent intended variables. Results showed that parameter estimates can differ substantially across different levels of proxy variable representation. When imperfect proxy variables are used, an insufficient amount of variance is removed from the observed spurious relationship between design variables. The findings from this methodological demonstration underscore the importance of precisely imbuing confounds with conceptual meaning and selecting proxy variables that accurately represent the underlying construct for which control is intended.

Current deficits in the rehabilitation psychology literature involving longitudinal studies investigating positive outcomes following acquired disabilities have deserved research attention. In the current study, data on happiness as an enduring mood tone, as measured by the Life Satisfaction Index (LSI) was collected from 1271 individuals (“insiders”) having incurred either a traumatic brain injury (TBI), spinal cord injury (SCI), severe burn, or intra-articular fracture (IAF) or from someone who felt close enough to speak on their behalf (“outsiders”). Data on happiness, functional independence as measured by the Functional Independence Measure (FIM), and other variables of interest were collected at 12 months, 24 months, 48 months, and 60 months after being medically discharged. Hierarchical Linear Modeling (HLM) analyses showed that trajectories of happiness remained stable across participants and did not change significantly over five years post-discharge regardless of injury type, FIM, or insider/outsider status. Happiness was significantly predicted by FIM, injury type, and whether the respondent was an insider or outsider. Those who were more impaired and less functionally independent were less happy. Those with a TBI were consistently less happy than those with an IAF or SCI and outsiders reported greater happiness on behalf of the insider than did the insiders themselves. This study shows that there is stability in happiness levels that can be sustained at least five years post-discharge and that there are discrepancies between insider and outsider reports of subjective happiness. Proxy reports can be used as valuable and valid secondary sources of information but should not be used as substitutes for first hand reports unless absolutely necessary.

Web pages take noticeably longer to load when accessing the Internet using high-latency wide-area wireless networks like 3G. This delay can result in lower user satisfaction and lost revenue for web site operators. By locating a just-in-time prefetching push proxy in the Internet service provider's mobile network core and routing mobile client web requests through it, web page load times can be perceivably reduced. Our analysis and experimental results demonstrate that the use of a push proxy results in a much smaller dependency on the mobile-client-to-network latency than seen in environments where no proxy is used; in particular, only one full round trip from client to server is necessary regardless of the number of resources referenced by a web page. In addition, we find that the ideal location for a push proxy is close to the servers that the mobile client accesses, minimizing the latency between the proxy and the servers that the mobile client accesses through it; this is in contrast to traditional prefetching proxies that do not push prefetched items to the client, which are best deployed halfway between the client and the server.

Zugl.: Diplomarbeit.

Techn. University, Diss., 2004--Berlin.

La Inteligencia de Negocios (BI) y la Minería de Datos (DM) son áreas que han tenido un importante crecimiento en los últimos diez años. Estas disciplinas ayudan en la toma de decisiones de las compañías, entregando valiosa información extraída desde los datos de la empresa mediante complejos procedimientos. Resulta indispensable para los proveedores de servicios en estas áreas el contar con los datos disponibles la mayor cantidad de tiempo posible, para no incurrir en quiebres importantes de su continuidad operacional. El objetivo de la presente memoria fue crear un sistema básico de alta disponibilidad que incluye replicación y distribución de consultas a Bases de Datos, el cual pretende disminuir los riesgos asociados al downtime de los servidores de datos de la empresa Penta Analytics. Esta compañía utiliza Infobright como motor de almacenamiento de datos, el cual está orientado a Bases de Datos de tipo analítica y que actualmente no cuenta con una solución de alta disponibilidad que abarque ambos temas. La solución fue diseñada en base a tres componentes principales: un sistema de replicación de datos, un sistema de distribución de consultas y una tabla de estados de replicación. Para la replicación se desarrolló una aplicación ad-hoc programada en lenguaje Java, mientras que la distribución de consultas fue creada en base a una aplicación llamada MySQL-Proxy, que fue adaptada para funcionar con un clúster de Bases de Datos analíticas. La tabla de replicación resguarda el estado de actualización de cada componente del sistema, estado que debe ser consistente ante cualquier escenario. Como resultado se obtuvo un sistema que mejora los tiempos de respaldo de datos y de respuestas a consultas a Bases de Datos. La solución para la distribución de consultas es escalable y paralelizable, mientras que el sistema de replicación escala sólo verticalmente y requiere modificación de código para agregar paralelismo y escalabilidad horizontal. Este desarrollo constituye una innovación como solución de alta disponibilidad para este tipo de Bases de Datos. En base a la presente memoria se muestra un conjunto de posibles mejoras y trabajos futuros, como por ejemplo mejorar los algoritmos de selección de servidor en la distribución de consultas u optimizar el transporte de datos entre servidores del sistema de replicación.

This study examines whether firms' decisions to offer company stock in defined contribution (DC) plans are explained by managers' corporate control motives. Using a large sample of proxy voting outcomes, I find that employee ownership in DC plans is significantly and positively associated with the level of voting support for management sponsored proposals. This suggests that managers encourage employee DC holdings in company stock in order to receive higher voting support in favor of management. The effects of employee ownership on voting outcomes are significantly greater in subsample tests than in full sample tests: management proposals opposed by Institutional Shareholder Services, management proposals of close votes, director election votes receiving more than 20% of votes withheld, and say-on-pay frequency proposals.

Context. Many applications today use databases to store user informationor other data for their applications. This information can beaccessed through various different languages depending on what typeof database it is. Databases that use SQL can maliciously be exploitedwith SQL injection attacks. This type of attack involves inserting SQLcode in the query parameter. The injected code sent from the clientwill then be executed on the database. This can lead to unauthorizedaccess to data or other modifications within the database. Objectives. In this study we investigate if a system can be builtwhich prevents SQL injection attacks from succeeding on web applicationsthat is connected with a MySQL database. In the intendedmodel, a proxy is placed between the web server and the database.The purpose of the proxy is to hash the SQL query parameter dataand remove any characters that the database will interpret as commentsyntax. By processing each query before it reaches its destination webelieve we can prevent vulnerable SQL injection points from being exploited. Methods. A literary study is conducted the gain the knowledgeneeded to accomplish the objectives for this thesis. A proxy is developedand tested within a system containing a web server and database.The tests are analyzed to arrive at a conclusion that answers ours researchquestions. Results. Six tests are conducted which includes detection of vulnerableSQL injection points and the delay difference on the system withand without the proxy. The result is presented and analyzed in thethesis. Conclusions. We conclude that the proxy prevents SQL injectionpoints to be vulnerable on the web application. Vulnerable SQL injectionpoints is still reported even with the proxy deployed in thesystem. The web server is able to process more http requests that requiresa database query when the proxy is not used within the system.More studies are required since there is still vulnerable SQL injectionspoints.

The problem of SSL/TLS interception ("trusted proxy in SSL/TLS connection") has been known for years and many implementations exist. However, all of them share a single technical solution which is based solely on the PKI authentication mechanism and suffers from multiple serious disadvantages. Most importantly, it is not compatible with several aspects or future trends of SSL/TLS and PKI, there's almost no space for improvement and its real use may spawn legal issues. After we analyze technical background and the current solution, we will propose another one, based not only on PKI but SSL/TLS too. Both solutions will be compared and general superiority of the new one will be shown. Basic implementation and analysis will follow, along with deployment requirements and ideas for future development. Powered by TCPDF (www.tcpdf.org)

Despite recommendations, health services interventions often lack mechanistic evidence. One way of acquiring such evidence is through the use of hypothetical settings. However, there are knowledge gaps about the factors that affect whether hypothetical decisions will correlate to real-world ones. This thesis involves a systematic concept review to identify the factors important to the hypothetical-real-world decision relationship and a case study of the relevance of these factors to the decision aid development context. The 42 identified factors were grouped into 5 categories: Personal Characteristics, Presentation Characteristics, Cognitive Factors, Motivation, and Participant Characteristics. Through cognitive interviews with decision aid developers we found that the interviewees had generally not considered many of the factors thought to affect the external validity of hypothetical data. This thesis contributes a framework that highlights factors for investigators to consider when designing studies in hypothetical settings and that can serve as a foundation for future research.