行为学、社会学、心理学和医药学方面，结构方程模型(SEMs) 是研究有关潜在变量最常用的模型。这篇论文的目的是研究基本和高级结构方程模型的贝叶斯诊断，本文研究的结构方程模型包括非线性纺构方程模型、变换结构方程模型、二层结构方程模型和混合结构方程模型。基于对数贝叶斯因子的一阶与二阶局部影响测度是本文进行贝贝叶斯诊断的基础。局部影响测度的计算和模型参数估计是利用了蒙特卡洛(MCMC) 和扩展数据的方法。对比传统的基于极大似然的诊断，本文提出的贝叶斯诊断方法不仅能检测异常点或者影响点，而且可以诊断模型假设和先验设定的敏感性。 这些是通过对数据、模型假设和先验设定进行不同的扰动获得的 本文用大量的模拟实验来说明所提出的贝叶斯诊断方法的作用。 本文基于不同类型的结构方程模型，应用所提出的贝叶斯诊断方法于一些实际数据。 / In the behavioral, social, psychological, and medical sciences, the most widely used models in assessing latent variables are structural equation models (SEMs). This thesis aims to develop Bayesian diagnostic procedures for basic and advanced SEMs such as nonlinear SEMs, transformation SEMs, two-level SEMs, and mixture SEMs. The first- and second-order local inference measures with the objective functions defined based on the logarithm of Bayes factor are proposed to perform the Bayesian diagnostics. Markov chain Monte Carlo (MCMC) methods, along with data augmentation, are developed to compute the local influence measures and to estimate unknown model parameters. Compared with conventional maximum likelihood-based diagnostic procedures, the proposed Bayesian diagnostic approach can not only detect outliers or influential points in the observed data, but also conduct model comparison and sensitivity analysis by perturbing the data, sampling distributions, and the prior distributions of model parameters via a variety of perturbations. The empirical performances of the proposed Bayesian diagnostic procedures are revealed through extensive simulation studies. Several real-life data sets are used to illustrate the application of our proposed methodology in the context of different SEMs. / Detailed summary in vernacular field only. / Chen, Ji. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2013. / Includes bibliographical references (leaves 130-135). / Abstract also in Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Structural equation models --- p.1 / Chapter 1.2 --- Bayesian diagnostics --- p.3 / Chapter 1.2.1 --- The first and second order local influence measures --- p.5 / Chapter 1.2.2 --- A simple example --- p.9 / Chapter 2 --- Bayesian diagnostics of nonlinear SEMs --- p.15 / Chapter 2.1 --- Model description --- p.16 / Chapter 2.2 --- Bayesian estimation and local inference of nonlinear SEMs --- p.17 / Chapter 2.3 --- Simulation study --- p.24 / Chapter 2.3.1 --- Simulation study 1 --- p.24 / Chapter 2.3.2 --- Simulation study 2 --- p.25 / Chapter 2.3.3 --- Simulation study 3 --- p.27 / Chapter 2.4 --- Application: A study of kidney disease for type 2 diabetic patients --- p.29 / Chapter 3 --- Bayesian diagnostics of transformation SEMs --- p.40 / Chapter 3.1 --- Model description --- p.41 / Chapter 3.2 --- Bayesian estimation and local inference of the transformation SEMs --- p.44 / Chapter 3.3 --- Simulation study --- p.54 / Chapter 3.3.1 --- Simulation study 1 --- p.54 / Chapter 3.3.2 --- Simulation study 2 --- p.56 / Chapter 3.4 --- Application: A study on the risk factors of osteoporotic fracture in older people --- p.58 / Chapter 4 --- Bayesian diagnostics of two-level SEMs --- p.73 / Chapter 4.1 --- Model description --- p.74 / Chapter 4.2 --- Bayesian estimation and local inference of two-level SEMs --- p.75 / Chapter 4.3 --- Simulation study --- p.88 / Chapter 4.4 --- Application: A study of AIDS data --- p.91 / Chapter 5 --- Bayesian diagnostics of mixture SEMs --- p.106 / Chapter 5.1 --- Model description --- p.107 / Chapter 5.2 --- Bayesian estimation and local inference ofmixture SEMs --- p.108 / Chapter 5.3 --- Simulation study --- p.116 / Chapter 5.3.1 --- Simulation study 1 --- p.116 / Chapter 5.3.2 --- Simulation study 2 --- p.118 / Chapter 6 --- Conclusion --- p.126 / Bibliography --- p.130 / Chapter A --- Proof of Theorem 1.1 and 1.2 --- p.136 / Chapter B --- Full conditional distributions of the nonlinear SEM --- p.138 / Chapter C --- Full conditional distributions of the transformation SEM --- p.141 / Chapter D --- Full conditional distributions of the two-level SEM --- p.144 / Chapter E --- AIDS preventative intervention data --- p.150 / Chapter F --- Permutation sampler in the mixture SEM --- p.152 / Chapter G --- Full conditional distributions of the mixture SEM --- p.153

In practice, the application and extension of the ICA model depend on the problem and the data to be investigated. We finally focus on GARCH models in finance, and show that estimation of univariate or multivariate GARCH models is actually a nonlinear ICA problem; maximizing the likelihood is equivalent to minimizing the statistical dependence in standardized residuals. ICA can then be used for factor extraction in multivariate factor GARCH models. We also develop some extensions of ICA for this task. These techniques for extracting factors from multivariate return series are compared both theoretically and experimentally. We find that the one based on conditional decorrelation between factors behaves best. / In this thesis, first we consider the problem of source separation of post-nonlinear (PNL) mixtures, which is an extension of ICA to the nonlinear mixing case. With a large number of parameters, existing methods are computation-demanding and may be prone to local optima. Based on the fact that linear mixtures of independent variables tend to be Gaussian, we develop a simple and efficient method for this problem, namely extended Gaussianization. With Gaussianization as preprocessing, this method approximates each linear mixture of independent sources by the Cornish-Fisher expansion with only two parameters. Inspired by the relationship between the PNL mixing model and the Wiener system, extended Gaussianization is also proposed for blind inversion of Wiener systems. / Independent component analysis (ICA) is a recent and powerful technique for recovering latent independent sources given only their mixtures. The basic ICA model assumes that sources are linearly mixed and mutually independent. / Next, we study the subband decomposition ICA (SDICA) model, which extends the basic ICA model to allow dependence between sources by assuming that only some narrow-band source sub-components are independent. In SDICA, it is difficult to determine the subbands of source independent sub-components. We discuss the feasibility of performing SDICA in an adaptive manner. An adaptive method, called band selective ICA, is then proposed for this task. We also investigate the relationship between overcomplete ICA and SDICA and show that band selective ICA can solve the overcomplete ICA problems with sources having specific frequency localizations. Experimental results on separating images of human faces as well as artificial data are presented to verify the powerfulness of band selective ICA. / Zhang Kun. / "July 2005." / Adviser: Lai-Wan Chan. / Source: Dissertation Abstracts International, Volume: 67-07, Section: B, page: 3925. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2005. / Includes bibliographical references (p. 218-234). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract in English and Chinese. / School code: 1307.

Latent variable models (LVMS) are widely appreciated multivariate methods to explore variables that are related to the observed variables, and assessing the relationships among them. One of most widely used latent variable models is structural equation model (SEM). Based on more than a dozen standard packages for fitting SEMs, such as LISREL VIII (Jorskog and Sorbom, 1996), and EQS (Bentler, 2004), these models have been widely appreciated in behavioral, educational, medical, social, and psychological research. The statistical theories and methods in these packages are based on the normal distribution; hence, they are vulnerable to outliers and the non-normal assumption. As outliers and non-normal data set are commonly encountered in substantive research, this fundamental problem has received much attention in the field. However, almost all existing methods are developed via the covariance structure analysis approach that heavily depends on the asymptotical properties of the sample covariance matrices S. Hence, this approach cannot be applied to the more complex SEMs and/or SEMs with more complex data structure such as missing data, because under these more complicated situations S is complicated, and its asymptotical properties are not well known. The objectives of this thesis are to develop novel robust methods for analyzing complex SEMs and/or more data structures, including but not limited to nonlinear SEMs with missing data. Both maximum likelihood (ML) and Bayesian approaches for estimation, hypothesis testing and model comparison will be investigated. Efficient algorithm for computing the results for statistical inference will be developed through unitization and modification of the advanced tools in statistical computing, for example the Monte Carlo Expectation-Maximization algorithm, and the Markov Chains Monte Carlo methods. Asymptotical properties of some statistics are derived. Simulation studies and real examples are conducted to reveal the empirical performance of the Bayesian and ML approaches. The newly developed methodologies will be very useful for analyzing complex data in the substantive research. / Xia Yemao. / "October 2005." / Adviser: S. Y. Lee. / Source: Dissertation Abstracts International, Volume: 67-07, Section: B, page: 3883. / Thesis (Ph.D.)--Chinese University of Hong Kong, 2005. / Includes bibliographical references (p. 105-114). / Electronic reproduction. Hong Kong : Chinese University of Hong Kong, [2012] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Electronic reproduction. [Ann Arbor, MI] : ProQuest Information and Learning, [200-] System requirements: Adobe Acrobat Reader. Available via World Wide Web. / Abstract in English and Chinese. / School code: 1307.

by Wai-Yin Chan. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1996. / Includes bibliographical references (leaves 42-45). / Chapter Chapter 1 --- Introduction / Chapter 1.1 --- Main Problems --- p.1 / Chapter 1.2 --- Class of Regularized Estimator --- p.4 / Chapter 1.3 --- Preliminaries --- p.6 / Chapter 1.4 --- Related Works --- p.9 / Chapter 1.5 --- Brief Summary --- p.10 / Chapter Chapter 2 --- Estimation of the Covariance Matrix and its Eigenvalues in a Wishart Distribution / Chapter 2.1 --- Significance of The Problem --- p.12 / Chapter 2.2 --- Review of the Previous Work --- p.13 / Chapter 2.3 --- Properties of the Wishart Distribution --- p.18 / Chapter 2.4 --- Main Results --- p.20 / Chapter 2.5 --- Simulation Study --- p.23 / Chapter Chapter 3 --- Estimation of the Scale Matrix and its Eigenvalues in a Multivariate F Distribution / Chapter 3.1 --- Formulation and significance of the Problem --- p.26 / Chapter 3.2 --- Review of the Previous Works --- p.28 / Chapter 3.3 --- Properties of Multivariate F Distribution --- p.30 / Chapter 3.4 --- Main Results --- p.33 / Chapter 3.5 --- Simulation Study --- p.38 / Chapter Chapter 4 --- Further research --- p.40 / Reference --- p.42 / Appendix --- p.46

Chung Po-Yi. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2003. / Includes bibliographical references (leaves 33-34). / Abstracts in English and Chinese. / Chapter 1 --- Introduction --- p.1 / Chapter 2 --- Bayesian Approach --- p.4 / Chapter 2.1 --- Model Description --- p.5 / Chapter 2.2 --- Identification --- p.6 / Chapter 2.3 --- Bayesian Analysis of the Model --- p.8 / Chapter 2.3.1 --- Posterior Analysis --- p.8 / Chapter 2.3.2 --- The Gibbs Sampler --- p.9 / Chapter 2.3.3 --- Conditional Distributions --- p.10 / Chapter 2.4 --- Bayesian Estimation --- p.13 / Chapter 3 --- Two-stage Partition Approach --- p.15 / Chapter 3.1 --- First Stage: PRELIS --- p.15 / Chapter 3.2 --- Second Stage: LISREL --- p.17 / Chapter 3.2.1 --- Model Description --- p.17 / Chapter 3.2.2 --- Identification --- p.17 / Chapter 3.2.3 --- LISREL Analysis of the Model --- p.18 / Chapter 4 --- Comparison --- p.19 / Chapter 4.1 --- Simulation Studies --- p.19 / Chapter 4.2 --- Real Data Studies --- p.28 / Chapter 5 --- Conclusion & Discussion --- p.30 / Chapter A --- Tables for the Two Approaches --- p.35 / Chapter B --- Manifest variables in the ICPSR examples --- p.51 / Chapter C --- PRELIS & LISREL Scripts for Simulation Studies --- p.52

INTRODUÇÃO: A neutropenia febril (NF) é uma complicação freqüente e potencialmente fatal no manejo do paciente onco-hematológico. Estudos recentes demonstram que a NF consiste em um grupo heterogêneo de pacientes com riscos variados. Nosso objetivo foi avaliar a taxa de falência ao tratamento de primeira linha, taxa de internação e óbito. Além disso, estudamos as variáveis clínico-laboratoriais em relação aos desfechos, a validação do índice Multinational Association for Supportive Care of Cancer (MASCC) modificado e a taxa de positividade de hemocultura e urocultura, como também o perfil de sensibilidade ao cefepima. CASUÍSTICA E METODOLOGIA: Estudo retrospectivo unicêntrico. Os dados foram obtidos através dos prontuários do Hospital-Dia no período de Julho de 2001 a Junho de 2006. Foram avaliados eventos com NF tratados com cefepima 2g (2x/dia), associado ou não, a teicoplamina 400mg/dia. RESULTADOS: Em 128 pacientes, estudamos 178 eventos de NF. A taxa de falência ao tratamento de primeira linha foi de 36,5%, taxa de internação 20,7% e óbito em 6,2% entre os eventos de NF. Na análise multivariada do estudo das categorias clínico-laboratoriais e dos desfechos encontramos como dados significantes em relação ao risco da falência ao tratamento de primeira linha: Idade < 60 anos (OR: 2,11 IC95%: 1,71-2,51, p = 0,004) e creatinina sérica > 1,2mg/dL (OR: 7,19, IC95%: 1,81 30,71 p= 0,005). Os dados significantes para o risco de internação foram: Ausência do diagnóstico de Linfoma não - Hodgkin (OR: 2,42 IC95%: 2,04 2,8, p= 0,011) Tabagismo (OR: 3,14, IC95% 1,14 8,66, p=0,027) e creatinina sérica > 1,2mg/dL (OR: 7,97, IC95% 21,19 - 28,95, p=0,002). Em relação ao óbito, o único dado de risco significante foi a saturação de oxigênio < 95% (OR: 5,8, IC95% 1,50 - 22,56, p = 0,011). Em relação ao índice MASCC modificado e seu impacto sobre os desfechos obtivemos os seguintes resultados: Falência do tratamento de primeira linha e (baixo risco versus alto risco): 35,2% x 53,8%, p=0,232; Internação (baixo risco versus alto risco): 18,2% x 53,8%, p = 0,006; óbito (baixo risco versus alto risco): 4,3% x 30,8%, p=0,004. As taxas de hemocultura e urocultura positivas foram respectivamente: 13% e 8%. O agente isolado mais freqüente nos dois exames foi Eschericia coli. Em relação ao perfil de sensibilidade dos agentes isolados e testados, 100% foram sensíveis ao cefepima. CONCLUSÕES: Os eventos de NF em tratamento ambulatorial apresentaram taxas satisfatórias em relação aos desfechos. Os dados sugerem que os riscos como: Ausência de Linfoma não - Hodgkin, tabagismo, creatinina sérica > 1,2mg/dL e oximetria de pulso < 95% merecem ser considerados como fatores de riscos para desfechos indesejáveis. O índice MASCC modificado mostrou-se eficaz para classificar os eventos classificados como alto risco na nossa população. Em relação aos agentes isolados e testados, 100% são sensíveis ao antibiótico de primeira linha cefepima. / BACKGROUND AND OBJECTIVES: Febrile Neutropenia (FN) is a frequent adverse event and potentially lethal in patients with haematologic malignancies. Nowadays, FN represents a heterogeneous group with different risk for serious complications and death. We studied the first line antibiotic failure, hospitalization rate and death. In addition, it was compared clinical and laboratory data with outcomes, validation of the usefulness of Modified Multinational Association for Supportive Care of Cancer (MASCC) and blood culture and urine culture rate identification. DESIGN AND METHODS: We elaborated a retrospective study. It was evaluated patients with haematologic malignancies who were treated with Cefepime 2g intravenous (IV) twice a day, with or without Teicoplanin 400mg (IV) once a day. RESULTS: Of the 178 FN events, it was observed: first line antibiotic failure 36,5%, hospitalization rate 20,7% and deaths 6,2%. In multivariate analyses, it was evidenced with risk to first line antibiotic failure: Age < 60 years (OR: 2,11, CI95%: 1,71-2,51, p =0,004), serum creatinine > 1,2mg/mL (OR: 7,19, CI95%: 1,81 30,71 p= 0,005). In hospitalization the risks were: Without diagnosis of Non- Hodgkin Lymphoma (OR: 2,42, CI95%: 2,04 2,8, p= 0,011), smoking (OR: 3,14, CI95% 1,14 8,66, p=0,027), serum creatinine > 1,2mg/dL (OR: 7,97, CI95%21,19- 28,95, p=0,002). Relating to death, the risk was transcutaneous oximetry < 95% (OR: 5.8, CI95%: 1.50 22.56, p = 0.011). Analyzing MASCC index, 165 events were classified as low-risk and 13 as high-risk. Outpatient treatment failures were reported in connection with 7 (53.8%) high-risk episodes and 30 (18.2%) low-risk, p=0.006. In addition, death in 7 (4.2%) low-risk and 4 (30.8%) high-risk events, p=0.004. Microbiological infection documented was identified in 13% and 8% in blood cultures and urine cultures, respectively. The most common agent isolated was E. coli and 100% were sensitive to cefepime. INTERPRETATIONS AND CONCLUSIONS: The outpatient treatment with intravenous antibiotic was satisfactory. The risks: Haematologic malignancies other than Non-Hodgkin Lymphoma, smoking, serum creatinine elevated and oximetry < 95% should be considered in FN evaluation. It was validated MASCC index in the Brazilian population. Relating to microbiological agents studied 100% were not resistant for cefepime.

Osteoarthritis (OA) is the most common joint disease in human populations with onset and severity influenced by mechanical loading, aging effects, genetics, anatomy, and body mass. Despite major advancements in knowledge, the aetiopathogenesis of OA is complex and still poorly understood. Lack of standardization in methods to quantify skeletal OA make it difficult to study the effects of interacting explanatory variables on arthritic response, and prevents comparison of results between bioarchaeological studies. Joint changes of OA as a function of both the natural aging process and of mechanical stress can make an individual appear older than their chronological age, potentially impacting current methods to derive accurate skeletal age at death estimates, particularly in load-bearing regions. This project addressed these issues through three studies, using a large skeletal sample of modern Europeans for which sex, age, and occupation were available. The first study used principal component analysis (PCA) as a standardized procedure to compute aggregate scores for joint complexes and a systemic measure of OA in each region of the lumbar spine, pelvis, and knee. The second study analyzed the composite scores with a multiple regression model to determine the relative contribution of three predictors: age, activity, and body size, and their effect on skeletal expression of OA in each region. Body size (stature and mass) was calculated from postcranial skeletal measurements; torsional strength (J) of the femoral midshaft was calculated from three-dimensional surface models, size standardized and used as a proxy for measure of activity. The third study considered the effect of OA severity on the validity and reliability of three methods to estimate age at death from load-bearing joints of the os coxa: the pubic symphysis, auricular surface, and acetabulum. The study was designed to determine whether OA in adults acts as a potential limitation or benefit in deriving accurate skeletal age at death estimates from pelvic joint morphology that will contribute to standardized methods in establishing physiological degeneration of the skeleton due to aging. Body size and activity factors did not contribute significantly to OA pathology outside of the age-related expression in either of the lumbar vertebrae or knee regions, and only demonstrated a weak association at pelvic joints. Differences in adult patterns of age are reflected in joint arthritic changes of the os coxa and OA severity has an effect on the accuracy of age estimates from the pelvis; those with OA consistently aging faster in all three joint areas. This influence is most significant for young individuals at the auricular surface and pubic symphysis, over-aging at both. Oldest persons with little arthritic patterning at the acetabulum were under-aged, but accuracy of the age estimate improved as OA severity increased. Systemic measures of OA determined through PCA as an indicator of age, appear useful to identify the very old, but may also help to distinguish between systemic age-related stresses and localized biomechanical effects. Interpreting OA as evidence for old age, measures of habitual activity, and larger body mass should be exercised with caution in skeletal populations. / Graduate / 2018-04-18 / 0327 / 0339 / 0571 / calce.stephanie@gmail.com

No description available.

Submitted by Jean Medeiros (jeanletras@uepb.edu.br) on 2016-05-04T14:15:29Z No. of bitstreams: 1 PDF - Ruana Chagas da Silva.pdf: 1951311 bytes, checksum: ecb7f6c829485e06a3d3fdd70af43a7a (MD5) / Approved for entry into archive by Secta BC (secta.csu.bc@uepb.edu.br) on 2016-07-25T19:31:11Z (GMT) No. of bitstreams: 1 PDF - Ruana Chagas da Silva.pdf: 1951311 bytes, checksum: ecb7f6c829485e06a3d3fdd70af43a7a (MD5) / Approved for entry into archive by Secta BC (secta.csu.bc@uepb.edu.br) on 2016-07-25T19:33:41Z (GMT) No. of bitstreams: 1 PDF - Ruana Chagas da Silva.pdf: 1951311 bytes, checksum: ecb7f6c829485e06a3d3fdd70af43a7a (MD5) / Made available in DSpace on 2016-07-25T19:33:41Z (GMT). No. of bitstreams: 1 PDF - Ruana Chagas da Silva.pdf: 1951311 bytes, checksum: ecb7f6c829485e06a3d3fdd70af43a7a (MD5) Previous issue date: 2016-03-08 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / This study aims to evaluate the effects of genotype x environment interaction by checking adaptability and stability, in addition to the grouping of environments through the AMMI methods (Additive Main effects and Multiplicative Interaction Analysis), GGE Biplot (Genotype and genotypes by Environment Interaction). The data were collected from the breeding program at Embrapa Cotton , referring to 16 cotton genotypes in eight locations (enviroment) in the state of Mato Grosso, Brazil. The analysis of individual and joint variance were carried out, then the analysis of adaptability and stability through the AMMI and GGE Biplot methods. Regarding the proportion of the capture of GE interaction in the first two axes, according to the two methodologies, a small superiority was noted of GGE (72.64%) compared to AMMI analysis (68.28%). The A4 and A6 environments were the main contributors to GE interaction, while the A2 and A3 environment the least influenced by GE interaction in both methods. The G4 genotypes (FMT 701) and G9 (LD CV 05) stood out as the most stable, combining broad adaptation and productivity, followed by G12 (Nuopal), G14 (CNPA MT 04 2080, G15 (CNPA MT 04 in 2088) and G16 (CNPA GO 03 -1947). The GGE methodology proved to be easier to interpret as well as less subjective to provide results than AMMI. / O presente estudo foi realizado com o objetivo de avaliar os efeitos da interação genótipo x ambiente, analisar adaptabilidade e estabilidade dos genótipos, além do agrupamento dos ambientes, por meio dos métodos AMMI (Additive Main effects and Multiplicative Interaction Analysis), GGE Biplot (Genotype and Genotypes by Environment Interaction). Foram utilizados dados obtidos junto ao programa de melhoramento genético da Embrapa Algodão, referentes a 16 genótipos de algodão, em oito locais (ambientes) distribuídos no Mato Grosso. Foram realizadas analises de variância individual e conjunta, em seguida, as analises de adaptabilidade e estabilidade por meio dos métodos AMMI e GGE Biplot. Em relação à proporção da captura da interação GxA nos dois primeiros eixos, pelas duas metodologias, notou-se uma pequena superioridade do GGE (72,64%) em relação à análise AMMI (68,28%). Os ambientes A4 e A6 foram os que mais contribuíram para a interação GxA, enquanto os ambiente A2 e A3 os que menos influenciaram pelos dois métodos. Os genótipos G4 (FMT 701) e G9 (LD CV 05) destacaram-se como mais estáveis, combinando ampla adaptação e produtividade, seguidos por G12 (Nuopal), G14 (CNPA MT 04 2080, G15 (CNPA MT 04 2088) e G16 (CNPA GO 03 -1947). A metodologia GGE mostrou-se ser de mais fácil interpretação e com resultados menos subjetivos do que AMMI.

A avaliação da pós-graduação no Brasil realizada pela CAPES teve início em 1977. Atualmente, os resultados da avaliação da CAPES têm forte influência em alguns órgãos de alocação de recursos públicos. Há críticas da comunidade científica sobre o excesso de informações quantitativas avaliadas, e não se sabe ao certo qual a contribuição de cada variável na formação do conceito final. Este trabalho teve como objetivo principal medir a influência das variáveis quantitativas nos conceitos atribuídos pela CAPES aos programas de pós-graduação em Administração. Foram aplicadas as técnicas estatísticas multivariadas de análise discriminante e regressão logística ordinal. Com apenas quatro variáveis quantitativas, a análise discriminante classificou corretamente de 50% a 70% dos casos, sendo que as variáveis referentes à publicação científica tiveram maior influência nos resultados. A aplicação da regressão logística ordinal permitiu construir um instrumento de previsão de conceitos, que pode ser utilizado pelos gestores dos programas de pós-graduação em Administração. A teoria de regressão logística ordinal é apresentada de forma didática. Em termos metodológicos, esta é uma pesquisa quantitativa descritiva. A população foi formada pelos trinta e cinco programas de pós-graduação em Administração avaliados no triênio 2001-2003. / CAPES has been evaluating the postgraduate Brazilian system since 1977. Nowadays, these results are important to get public budgets. Scientific community complains about the excessive quantitative information that is required by actual evaluation rules. In fact, the contributions of each program?s feature to final grade are unknown. This research aimed to evaluate the contribution of quantitative variables to final grades of Management postgraduate programs. Discriminant analysis and ordinal logistic regression were applied. The classification procedure of discriminant analysis were successful every time, reaching between 50 and 70 percent of right classifications. Variables related to scientific publication are the most influence on final grade. A predict rule was made by ordinal logistic regression results. This is a useful tool to postgraduate programs managers. This work have a clearly text about ordinal logistic regression. This is a quantitative research. The population consisted of third five Management postgraduate programs evaluated from 2001 to 2003.