Indiana University-Purdue University Indianapolis (IUPUI) / The purpose of this study is to examine the informed consent (IC) procedure as it was experienced by patients who had undergone a de novo transjugular intrahepatic portosystemic shunt (TIPS), chemoembolization (TACE), or radioembolization (TARE) procedure in an Interventional Radiology (IR) Department. The three main study aims and a fourth exploratory aim are as follows: (1) Describe how patients who underwent a de novo TIPS, TACE, or TARE procedure in an IR Department described the IC procedure; (2) Describe what information patients who underwent a de novo TIPS, TACE, or TARE procedure in an IR Department recalled being told during the IC procedure; (3) Describe the satisfaction of patients who underwent a de novo TIPS, TACE, or TARE procedure in an IR Department with the IC procedure; and (4) Explore how the IC experiences of patients who underwent a de novo TIPS, TACE, or TARE procedure in an IR Department differed according to their levels of health literacy. Using a qualitative descriptive design, participants were recruited from an IR department that performed these procedures. A total of 14 participants were interviewed about their IC experiences and the Newest Vital Sign (NVS) Health Literacy assessment was administered. The participants described the IC procedure by discussing the staff they encountered, their feelings during the visit, the support persons who accompanied them, and the decisions they made about the procedure. The participants recalled being told about how their procedure would be performed, the care they would need, and the benefits and risks of the procedure. Most were satisfied with the information received during the IC procedure and found the information consistent with how they experienced the procedure. A few participants would have liked more visual materials, addition details about the procedure, simpler language, or more explanation of the medical terminology. No apparent differences in the IC experience could be attributed to health literacy. These findings suggest that persons’ experiences during the IC process are multi-faceted and affected by their emotions and concerns and the nature of their encounters with their healthcare providers.

Neste trabalho, por meio de experimentos simulados no delineamento inteiramente casualizado pelo método de Monte Carlo, inicialmente comparou-se o resultado do teste F global da análise de variância com o resultado de cada um dos seguintes procedimentos para a comparação de médias de tratamentos entre si duas a duas: Tukey, Duncan, DMS de Fisher, Student-Newman-Keuls e Scheffé. Ainda, os níveis de significância por comparação e por experimento, condicionais e incondicionais a um resultado preliminar significativo do teste F global, foram estimados para cada um dos testes de médias por meio do mesmo processo de simulação. Por fim, o poder condicional e incondicional de cada um dos procedimentos também foi estimado. Neste estudo notou-se a presença das duas situações contraditórias possíveis ao confrontar o resultado do teste F global com o resultado de cada um dos testes de médias analisados. No entanto, a presença dessas situações parece não estar ligada ao desvio de ao menos uma das pressuposições da análise. Notou-se, todavia, que testes que apresentaram maior incidência da situação de um teste F não significativo e ao menos um contraste entre médias significativo, apresentaram também estimativa para o nível de significância por experimento muito acima do nível nominal de 5% declarado. Além disso, testes com maior incidência da situação de um teste F global significativo e nenhum contraste entre médias significativo mostraram-se menos poderosos que os demais procedimentos avaliados. Verificou-se também que a aplicação condicional dos testes de comparações múltiplas a um resultado preliminar significativo do teste F global pode levar a um nível de significância real maior do que o nível nominal declarado. Essa aplicação condicional pode, ainda, levar a um pequeno aumento no poder dos testes em experimentos em que o número de tratamentos e a diferença real entre médias adjacentes são baixos. / In this work, through experiments simulated in completely randomized design by Monte Carlo\'s method, initially the result of the overall F-test on analysis of variance was compared with the result of each one of the following multiple comparison procedures: Tukey, Duncan, Fisher\'s least significant difference, Student-Newman-Keuls and Scheffé. Furthermore, the comparisonwise and experimentwise error rates, conditional and unconditional to a significant result of the overall F-test, were estimated for each one of the tests analyzed by the same simulation process. Finally, the conditional and unconditional power of each procedure was also estimated. In this study it was noticed the presence of the two possible conflicting situations to confront the overall F-test result with the result of each multiple comparison test analyzed. However, the presence of these situations doesn\'t seem to be linked to deviation at least one of the assumptions of the analysis. It was noted, however, tests that had a higher incidence of the situation with not significant F-test and at least one significant contrast between means, also showed estimate for experimentwise error rate much bigger than the nominal level of 5% declared. In addition, tests with higher incidence of the situation with significant overall F-test and no significant contrast between means, were less powerful than the other evaluated procedures. It was also found that the conditional application of multiple comparison tests to a significant result of overall F-test can lead to real significance level greater than the nominal level declared. This conditional application can also lead to a small increase in the power of tests in experiments in which the number of treatments and the real difference between adjacent means are small.

Neste trabalho, por meio de experimentos simulados no delineamento inteiramente casualizado pelo método de Monte Carlo, inicialmente comparou-se o resultado do teste F global da análise de variância com o resultado de cada um dos seguintes procedimentos para a comparação de médias de tratamentos entre si duas a duas: Tukey, Duncan, DMS de Fisher, Student-Newman-Keuls e Scheffé. Ainda, os níveis de significância por comparação e por experimento, condicionais e incondicionais a um resultado preliminar significativo do teste F global, foram estimados para cada um dos testes de médias por meio do mesmo processo de simulação. Por fim, o poder condicional e incondicional de cada um dos procedimentos também foi estimado. Neste estudo notou-se a presença das duas situações contraditórias possíveis ao confrontar o resultado do teste F global com o resultado de cada um dos testes de médias analisados. No entanto, a presença dessas situações parece não estar ligada ao desvio de ao menos uma das pressuposições da análise. Notou-se, todavia, que testes que apresentaram maior incidência da situação de um teste F não significativo e ao menos um contraste entre médias significativo, apresentaram também estimativa para o nível de significância por experimento muito acima do nível nominal de 5% declarado. Além disso, testes com maior incidência da situação de um teste F global significativo e nenhum contraste entre médias significativo mostraram-se menos poderosos que os demais procedimentos avaliados. Verificou-se também que a aplicação condicional dos testes de comparações múltiplas a um resultado preliminar significativo do teste F global pode levar a um nível de significância real maior do que o nível nominal declarado. Essa aplicação condicional pode, ainda, levar a um pequeno aumento no poder dos testes em experimentos em que o número de tratamentos e a diferença real entre médias adjacentes são baixos. / In this work, through experiments simulated in completely randomized design by Monte Carlo\'s method, initially the result of the overall F-test on analysis of variance was compared with the result of each one of the following multiple comparison procedures: Tukey, Duncan, Fisher\'s least significant difference, Student-Newman-Keuls and Scheffé. Furthermore, the comparisonwise and experimentwise error rates, conditional and unconditional to a significant result of the overall F-test, were estimated for each one of the tests analyzed by the same simulation process. Finally, the conditional and unconditional power of each procedure was also estimated. In this study it was noticed the presence of the two possible conflicting situations to confront the overall F-test result with the result of each multiple comparison test analyzed. However, the presence of these situations doesn\'t seem to be linked to deviation at least one of the assumptions of the analysis. It was noted, however, tests that had a higher incidence of the situation with not significant F-test and at least one significant contrast between means, also showed estimate for experimentwise error rate much bigger than the nominal level of 5% declared. In addition, tests with higher incidence of the situation with significant overall F-test and no significant contrast between means, were less powerful than the other evaluated procedures. It was also found that the conditional application of multiple comparison tests to a significant result of overall F-test can lead to real significance level greater than the nominal level declared. This conditional application can also lead to a small increase in the power of tests in experiments in which the number of treatments and the real difference between adjacent means are small.

In this thesis, techniques for testing aircraft models in ground effect in a low speed wind tunnel are investigated. Although these types of tests have been done before, the current study is unique in that forces are measured with an overhead balance instead of an internal balance. This has the advantage that the types of models that are difficult to mount on a sting with an internal balance, can often be mounted with a strut protruding from the top of the model. Positioning a sting-mounted model close to the ground at a high angle-of-attack is also usually difficult if not impossible. Finally, drag measurements are often more accurate when measured with an overhead balance rather than with an internal sting-type balance. The disadvantages associated with this method of testing are identified and solutions suggested. These include accurate moment transfers and correcting for support tares and interference. The thesis also investigates general procedures associated with ground effect testing such as proper boundary corrections and the necessity of a rolling floor. A simplified preliminary test series was performed in order to identify shortcomings in the existing equipment and procedures. This series is explained in Chapter 2. Chapter 4 and 5 describe changes made to the existing equipment following this test series. These include a novel telescopic fairing to shroud the mounting strut and an internal pitching mechanism. The correction techniques and general theory are summarised in Chapter 3. The author concludes in Chapter 6 that with the application of the techniques described in this thesis, the test engineer should be able to obtain accurate and reliable data from most aircraft configurations. Additional suggestions for testing models in ground effect are also given in this chapter. Finally, a few shortcomings that still need to be investigated are mentioned at the end of Chapter 6. AFRIKAANS : Hierdie verhandeling ondersoek tegnieke om vliegtuigmodelle in grondeffek in 'n laespoed-windtonnel te toets. Alhoewel hierdie tipe van toetse al voorheen gedoen is, is die huidige studie uniek deurdat 'n oorhoofse balans eerder as 'n interne balans gebruik word. Die voordeel hiervan is dat modelle wat moeilik op 'n naald- of "sting"-balans monteer kan word, baie keer makliker monteer kan word met 'n stang wat deur die bokant van die model steek. Posisioneering van 'n naald-gemonteerde model naby aan die vloer van die tonnel by hoe invalshoeke is gewoonlik ook baie moeilik indien nie onmoontlik nie. Laastens is sleurkrag-metings wat met 'n oorhoofse balans gemeet is gewoonlik meer akkuraat as sleurkrag-metings wat met 'n interne naald-tipe balans gedoen is. Die nadele wat met hierdie toetsmetode geassosieer kan word, word geïdentifiseer en moontlike oplossing word voorgestel. Hierdie sluit die berekening in van akkurate moment-transformasies en monteersleureffekte en -steurings. Die verhandeling ondersoek ook algemene prosedures wat met grondeffektoetse geassosieer kan word, byvoorbeeld akkurate wandkorreksies en die nodigheid van die rolvloer. 'n Vereenvoudigde vooraf-toetsreeks was uitgevoer om moontlike tekortkominge in die bestaande toerusting en prosedures te identifiseer. Hierdie toetsreeks word in Hoofstuk 2 bespreek. Hoofstuk 4 en 5 verduidelik die veranderinge wat aan die bestaande toerusting gemaak is na aanleidng van hierdie toetsreeks. Hierdie veranderinge sluit 'n teleskopiese windskerm in om die monteerstang te isoleer van die wind, sowel as 'n interne heimeganisme om die invalshoek van die model te verstel. Die korreksieprosedures en algemene teorie word in Hoofstuk 3 opgesom. Die outeur se gevolgtekking in Hoofstuk 6 stel dat die toetsingenieur, met behulp van die gebruik van die tegnieke in hierdie verhandeling beskryf, in staat behoort te wees om betroubare metings te kan neem van meeste vliegtuigkonfigurasies. Verdere voorstelle vir die toets van modelle in grondeffek word ook in hierdie hoofstuk gemaak. Uiteindelik word 'n paar tekortkominge genoem wat moontlik in 'n toekomstige studie ondersoek kan word. / Dissertation (MEng)--University of Pretoria, 1999. / Mechanical and Aeronautical Engineering / unrestricted