Because both validity and reliability indices are a function of the scores on a given administration of a measure, their values can often vary across samples. It is a common mistake to say that a test is reliable when in fact it is not the test that is reliable but the scores on the test that are reliable. In 1998, vacha-haase proposed a fixed-effects meta-analytic method for evaluating reliability that is similar to validity generalization studies called reliability generalization (rg). This study was conducted to evaluate alternative analysis strategies for the meta-analysis method of reliability generalization when the reliability estimates are not statistically independent. Five approaches for handling the violation of independence were implemented: ignoring the violation and treating each observation as independent, calculating one mean or median from each study, randomly selecting only one observation per study, or using a mixed effects model. This Monte Carlo study included five factors in the method. These factors were (a) the coefficient alpha, (b) sample size in the primary studies, (c) number of primary studies in the rg study, (d) number of reliability estimates from each, and (e) the degree of violation of independence where the strength of the dependence is related to the number of reliability indices (i.e. coefficient alpha) derived from a simulated set of examines and the magnitude of the correlation between the journal studies (with intra-class correlation icc = 0, .0l , .30, and .90). These factors were used to simulate samples under known and controlled population conditions. In general, the results suggested that the type of treatment does not have a noticeable impact on the accuracy of the reliability results but that researchers should be cautious when the intra-class correlation is relatively large. In addition, the simulations in this study resulted in very poor confidence band coverage. This research suggested that RG meta-analysis methods are appropriate for describing the overall average reliability of a measure or construct but the RG researcher should be careful in regards to the construction of confidence intervals.

Conselho Nacional de Desenvolvimento Científico e Tecnológico / CHAPTER II: The standardization of inter-laboratory results of germination test of forest species seeds requires that the methods be robust and subject to low variability. Therefore, the objective was to compare and discuss different forms of the coefficient of variation for normal seedlings in the method validation process for testing germination of seeds of 20 species of the Fabaceae family. Coefficients of variation for the experiment by lot and by laboratory were calculated for normal seedlings from the statistical analysis of method validations. For normal seedlings of 20 Brazilian forest species, the coefficients of variation were low (up to 9.84%), to average (up to 17.66%), contrary to expectations due to high genetic variability in these barely improved species. The increase of the coefficient is not related to treatment for breaking dormancy, but it grows as the lot quality decreases. The high coefficients by laboratory, overestimated by the lot effect, are uniform indicating that the methods are repeatable. The coefficient is not an indicator capable of predicting the heterogeneity of model variance. As normal distribution models random events, randomness is present in the validation process of the 20 forest species of the Fabaceae family. CHAPTER III: There are many treatments to overcome dormancy. However, descriptions of the consequences of these methods in seedling development are scarce. Because of the relevance of Fabaceae family in the context of dormancy seeds, seedlings and seeds of 10 forest species were evaluated quantitatively and qualitatively as to damage and infections caused by invasive treatments. Scarification, cutting and preheat methods were applied to seeds. They were then sampled was arranged in germitest, forming rolls distributed in a germination boxes under continuous fluorescent white light at 25 °C. Root protrusion used as the sole criterion, overestimates the efficiency of germination treatments to overcome dormancy. Scarification and cutting are efficient for Enterolobium contortisiliquum, Parkia pendula, Senna multijuga and Senna macranthera seeds, as well as cutting for Mimosa caesalpiniaefolia seeds. However, they are inefficient for Dimorphandra mollis and Enterolobium maximum because of the increase in the percentage of infected seedlings, and to Stryphnodendron adstringens, for dead seeds. The high percentage of damaged seedlings of Erythrina velutina and Erythrina speciosa cannot be attributed solely to treatment, because without pretreatment the root system gets strapped, forming a loop. The preheat treatment is inefficient because it results in a high percentage of hard seeds remaining. / CAPÍTULO II: A uniformização dos resultados inter laboratoriais de testes de germinação de sementes de espécies florestais exige que os métodos sejam robustos e sujeitos à baixa variabilidade. Assim, o objetivo foi comparar e discutir as diferentes formas de cálculo do coeficiente de variação para plântulas normais do processo de validação de métodos para teste de germinação de sementes de 20 espécies da família Fabaceae. Os experimentos de germinação de sementes para todas as espécies incluindo tratamentos pré-germinativos, tempos de contagem, substrato, temperatura e fotoperíodo foram validados com base no Manual de Validação da Associação Internacional para Teste de Sementes. Coeficientes de variação para o experimento, por lote e por laboratório para plântulas normais foram calculados, a partir das análises estatísticas previstas para a validação dos métodos, como exclusão de valores discrepantes, testes para as pressuposições do modelo e análise de variância. Para plântulas normais de 20 espécies florestais nativas, os coeficientes de variação foram de baixos (até 9,84%) a médios (até 17,66%), contrariando o esperado pela grande variabilidade genética dessas espécies pouco melhoradas. O aumento do coeficiente de variação não está relacionado ao tratamento de superação de dormência, porém cresce à medida que a qualidade do lote decresce. Os altos coeficientes de variação estimados por laboratório, superestimados pelo efeito de lotes, são uniformes indicando que os métodos são reproduzíveis. O coeficiente não é um indício capaz de predizer a heterogeneidade das variâncias e, consequentemente, a necessidade de transformação dos dados. Como a distribuição normal modela eventos aleatórios, a aleatoriedade está presente no processo de validação de métodos das 20 espécies da família Fabaceae, justificada pela distribuição normal dos resíduos. CAPÍTULO III: Muitos são os tratamentos descritos na literatura para superação da dormência de sementes, porém as consequências dos procedimentos no desenvolvimento das plântulas são raramente descritas. Pela relevância da família Fabaceae no contexto dos tratamentos para superação da dormência, plântulas e sementes de 10 espécies foram avaliadas quantitativamente e qualitativamente com o objetivo de determinar danos e infecções causados por tratamentos invasivos. Após revisão de literatura foram escolhidos métodos considerados eficientes para cada espécie. Experimentos foram conduzidos em delineamentos de blocos ao acaso e inteiramente casualizados, com sementes dispostas em papel germitest, com rolos dispostos em câmara de germinação sob luz branca fluorescente contínua a 25 ºC. Após a análise estatística selecionou-se o método que promoveu as maiores germinabilidades, maiores percentuais de plântulas normais, menores de anormais e de sementes mortas e duras, juntando-se outro relacionado ao mesmo procedimento e o tratamento térmico. A protrusão da raiz quando usado como critério único de germinação superestima a eficência dos tratamentos de superação da dormência de sementes de Fabaceae, porém é um indicador eficiente do potencial germinativo. A escarificação e o desponte são tratamentos eficientes de superação de dormência de sementes de Mimosa caesalpiniaefolia, Parkia pendula, Senna macranthera e Senna multijuga quando precedidos e seguidos pela assepsia com hipoclorito de sódio. Contudo, são ineficientes para sementes de Dimorphandra mollis e Enterolobium maximum pelo aumento dos percentuais de plântulas anormais infeccionadas e, para sementes de Stryphnodendron adstringens, pelo aumento dos percentuais de sementes mortas. Em Erythrina speciosa e Erythrina velutina os altos percentuais de plântulas anormais danificadas registrados a partir de sementes escarificadas e despontadas, não podem ser atribuídos exclusivamente aos tratamentos. A anormalidade causada pelo sistema radicular que fica preso no tegumento e enovela também é muito frequente em sementes sem qualquer pré-tratamento; O tratamento térmico úmido é ineficiente na superação de dormência de sementes de Fabaceae, mesmo com embebição posterior, pela desuniformidade pelos altos percentuais de sementes duras ao final do teste. / Mestre em Agronomia

Researchers perform multivariate techniques MANOVA, discriminant analysis and factor analysis. The most common applications in social science are to identify and test the effects from the analysis. The use of this multivariate technique is uncommon in investigating the effects of power usage and Province in South Africa on the amounts of the five power modes. This dissertation discusses this issue, the methodology and practical problems of the three multivariate techniques. The author examines the applications of each technique in social public research and comparisons are made between the three multivariate techniques. This dissertation concludes with a discussion of both the concepts of the present multivariate techniques and the results found on the use of the three multivariate techniques in the energy household consumption. The author recommends focusing on the hypotheses of the study or typical questions surrounding of each technique to guide the researcher in choosing the appropriate analysis in the social research, as each technique has some strengths and limitations. / Statistics / M. Sc. (Statistics)

Researchers perform multivariate techniques MANOVA, discriminant analysis and factor analysis. The most common applications in social science are to identify and test the effects from the analysis. The use of this multivariate technique is uncommon in investigating the effects of power usage and Province in South Africa on the amounts of the five power modes. This dissertation discusses this issue, the methodology and practical problems of the three multivariate techniques. The author examines the applications of each technique in social public research and comparisons are made between the three multivariate techniques. This dissertation concludes with a discussion of both the concepts of the present multivariate techniques and the results found on the use of the three multivariate techniques in the energy household consumption. The author recommends focusing on the hypotheses of the study or typical questions surrounding of each technique to guide the researcher in choosing the appropriate analysis in the social research, as each technique has some strengths and limitations. / Statistics / M. Sc. (Statistics)