O comportamento de liderança do treinador pode maximizar a performance do atleta ou até prejudicá-la, considerando os pontos de vista técnico e interpessoal. Uma das escalas mais utilizadas para avaliar comportamento de liderança é a LSS, proposta por Chelladurai e Saleh (1980). Porém, os estudos realizados com esta escala não exploraram as subdimensões intrínsecas destes comportamentos, além disso, verifica-se uma carência de escalas elaboradas no contexto brasileiro que considerem as idiossincrasias culturais e que avaliem, precisamente, e de forma independente, suas dimensões. Os objetivos da tese foram: propor e testar a validade de um novo instrumento de medida de liderança do treinador no contexto esportivo brasileiro, baseado no Modelo Multidimensional de Liderança no Esporte, identificando em cada uma das escalas de medida, da bateria que avalia o comportamento de liderança no esporte, a sua estrutura subjacente e qualidades psicométricas. Participaram do estudo 572 atletas, masculino = 410; feminino = 162, com idades variando de 13 a 35 anos ( = 15,79±3,17 anos), praticantes de modalidades individuais e coletivas. Dois instrumentos foram utilizados: um questionário Bio-Sócio-Demográfico e o segundo instrumento referente ao Comportamento de Liderança do Treinador, no qual cada dimensão foi composta por 10 itens formulados positivamente e com uma escala de concordância do tipo Likert, compreendida em 5 pontos, iniciando em “discordo fortemente” (1) até “concordo fortemente” (5). Cinco estudos foram conduzidos, testando separadamente cada dimensão: (1) treino e instrução, (2) comportamento democrático, (3) comportamento autocrático, (4) suporte social e (5) feedback positivo, no sentido de identificar quantos e quais são os fatores intrínsecos ou latentes em cada dimensão, verificando, também a adequação do modelo desenvolvido, inerente à escala de medida da dimensão aos dados disponíveis. Por fim, avaliou-se a confiabilidade dos resultados pelas medidas de precisão de cada uma das dimensões, em cada escala de medida. Os procedimentos estatísticos foram baseados em análises fatoriais exploratórias, confirmatória e modelagens de equação estrutural exploratória, assim como, cálculos de consistência interna, sendo todas as análises realizadas com base em matrizes policóricas. De modo geral, os resultados da bateria de testes para avaliar o comportamento de liderança do treinador evidenciaram variâncias explicativas das dimensões entre 35% e 57,12%, com índices de ajustamento absoluto, parcimonioso, comparativo e de correção considerados satisfatórios, demonstrando que os dados disponíveis se ajustam adequadamente aos modelos hipotéticos dos construtos avaliados. Os valores das medidas de precisão avaliadas foram satisfatórios, revelando que os itens em cada dimensão são consistentes entre si, representando uma medida precisa e de coerência interna do construto e de sua dimensão. Conclui-se o modelo de equação estrutural testado confirma, satisfatoriamente, a existência de cinco dimensões, com dois fatores de segunda ordem, consistentes e precisos na medida explicativa do Comportamento de Liderança do Treinador. / Coach's leadership behaviour can maximize the athlete's performance or even harm her, considering the technical and interpersonal viewpoints. One of the most widely used scales to evaluate leadership behaviour is LSS, proposed by Chelladurai and Saleh (1980). However, studies conducted with this scale have not explored the intrinsic subdimensions of these behaviors, in addition, there is a shortage of scales drawn up in the Brazilian context that consider our cultural idiosyncrasies and that they evaluate precisely and independently of their dimensions. The goals of the thesis were: propose and test the validity of a new trainer measurement instrument in the Brazilian sporting context, based on the multidimensional model of leadership in the sport, identifying in each of the measurement scales, the battery that evaluates leadership behavior in the sport, its underlying structure and psychometric qualities. Participated in the study 572 athletes, male = 410; Female = 162, with ages ranging from 13 to 35 years ( = 15,79±3,17), practitioners of individual and collective modalities. Two instruments were used: a bio-socio-demographic questionnaire and the second instrument for coach's leadership behaviour, in which each dimension was composed of 10 positively formulated items and with an agreement scale of type Likert, comprised of 5 points, starting "strongly disagree" (1) up to "strongly agree" (5). Five studies were conducted, testing separately each dimension: (1) training and instruction, (2) democratic behaviour, (3) autocratic behaviour, (4) social support and (5) positive feedback, in order to identify how many and what are intrinsic or latent factors in each dimension, checking, also the adequacy of the developed model, inherent to the scale of measurement of the dimension to the available data. Finally, the reliability of the results was evaluated by the precision measurements of each of the dimensions, on each measurement scale. The statistical procedures were based on exploratory factorials analyses, confirmatory and modelling structural equations, as well as internal consistency calculations, all analyses conducted based on policóricas matrices. Overall, the results of the test battery to evaluate the leadership behaviour of the trainer demonstrated the variances of the dimensions between 35% and 57.12%, with absolute, thrifty, compared and correction indices considered fairly satisfactory, demonstrating that the available data fits appropriately to the hypothetical models of the evaluated constructs. The values of the measured precision measurements were satisfactory, revealing that the items in each dimension are consistent with each other, representing an accurate measurement and internal coherence of the construct and its dimension. It is concluded that the model of structural equation tested satisfactorily confirms the existence of five dimensions, with two second-order factors, consistent and precise in the explanatory measure of the trainer's leadership behaviour.

A presente tese de doutorado trata da proposição de uma bateria de testes de medida de conteúdos pedagógicos (CP) no contexto de competições esportivas infantojuvenis. Foram considerados os CP do esporte mais citados pela literatura especializada: o desenvolvimento das habilidades motoras (HM), estratégico-tático (ET), socioeducativo (SE), de autonomia (Au), afetivo-social (AS) e democratização (De). A avaliação do quanto as competições esportivas contribuem para o desenvolvimento destes CP passa pela proposta de um novo instrumento de medida. Deste modo, esta tese teve como objetivo demonstrar as primeiras evidências de validade de uma bateria de testes de medida do favorecimento ao desenvolvimento de CP no esporte infantojuvenil, a qual é composta por seis escalas independentes, relativas a cada um dos CP: HM, ET, SE, Au, AS, De. Para tanto, no primeiro artigo, foi realizado uma revisão integrativa sobre estudos que apresentaram evidências de validade de instrumentos relacionados aos CP selecionados. Foi identificada a inexistência de modelos multidimensionais que avaliassem CP ou que tivessem a competição como elemento central da análise. No segundo artigo, foi elaborada a primeira versão da Bateria de Testes Gonçalves- Balbinotti de Favorecimento ao Desenvolvimento de Conteúdos Pedagógicos (BTGB-CP) e foram realizados cálculos de validade de conteúdo dos itens e da escala total para as dimensões Clareza de Linguagem (CL) e Pertinência Prática (PP), bem como medida a Dimensionalidade Teórica (DT) de cada item segundo a opinião de juízes-avaliadores bateria ficou reduzida a 59 itens com índices de validade de conteúdo (CVCCL = 0,942; CVCPP = 0,953) e concordância entre juízes (KDT = 0,822) satisfatórios. No terceiro artigo, foram estimadas as estruturas internas, testadas a estabilidade das estruturas internas e, por fim a fidedignidade de cada uma das seis escalas. Para tanto, foram executadas Modelagens de Equação Estrutural Exploratórias (ESEM) utilizando matrizes de correlação policóricas e foram computadas as consistências internas segundo os cálculos Alpha padronizado. Os resultados se mostraram satisfatórios conforme a estabilidade de suas respectivas estruturas e consistências internas. Por fim, no quarto artigo, foi proposto um novo modelo teórico-explicativo multidimensional, o qual foi testado a partir dos dados disponíveis. As correlações entre construtos demonstram a pertinência do modelo, a qual é suportada pelos índices de medida da estabilidade da estrutura do modelo segundo os dados disponíveis. Esta tese contribui, primeiramente, por demonstrar evidências de validade de seis escalas inéditas, as quais preenchem lacunas na avaliação de conteúdos pedagógicos. Ainda, contribui no ponto de vista teórico ao apresentar um detalhado modelo teórico-explicativo do construto geral relativo ao favorecimento ao desenvolvimento de CP no contexto esportivo infantojuvenil. / The present doctoral thesis deals with the proposition of a battery of measurement tests of pedagogical contents (PCs) in the context of youth sporting competitions. There were only considered the most evoked PCs of sports in the specialized literature: motor skills (MS), strategic/tactical (ST), socio-educative (SE), autonomy (Au), affective-social (AS) developments and democratization (De). The evaluation whether sporting competitions contribute to the development of these PCs comes after the proposition of a new measurement instrument. Thus, this thesis aimed to demonstrate the first evidences of validity of a battery of measurement tests of PCs’ development favoring in youth sports, which is composed by six independent scales, one for each PC: MS, ST , SE, Au, AS, De. Therefore, in the first article, an integrative review regarding studies, which presented evidences of validity of instruments related to the PCs, was carried out. It was identified a lack of multidimensional models which could evaluate PCs or that had sporting competitions as a central element of the analysis. In the second article, the first version of the Gonçalves-Balbinotti Battery of Tests of Pedagogical Contents’ Development Favoring (BTGB-CP) was elaborated and there were held the content validity calculations for each item and for the whole scale according to Language Clarity (LC), Practical Pertinence (PP) as well as Theoretical Dimensionality (TD) dimensions according to judges’ opinions. The battery was reduced to 59 items with satisfactory content validity (CVCLC = 0,942; CVCPP = 0,953) and judges agreement (KDT = 0,822) indices. In the third article, there were estimated the internal structures, tested the internal structures’ stability and the reliability of each one of the six scales Therefore, Exploratory Structural Equation Modeling (ESEM) using polychoric correlations were held and computed the internal consistency according to standardized Alpha. Results were satisfactory according to their respective internal structures and consistencies, indicating the relevance of the scales use. Lastly, in the fourth article, it was proposed a new theoretical-explanatory multidimensional model, which was tested based on the available data. Correlations between constructs demonstrate the model relevance, which is supported by the indices related to the model structure stability obtained with the available data. This thesis contributes, first, by demonstrating validity evidences of six new scales, which fill blanks regarding the evaluation of PCs. Furthermore, contributes from the theoretical point of view by presenting a detailed theoretical-explanatory model of the general construct related to PCs’ development favoring in the context of youth sports.

The aim of this study was to validate the Brandes´s Organizational Cynicism Scale in orderto use it with Argentinean workers. Functional equivalence of the instrument was analyzed. The adapted version was administered to a sample of 396 workers; who also answered a selection of instruments to explore organizational trust, organizational commitment and job satisfaction. Exploratory factor analysis showed a three-factor structure that explained 68.8% of the total variance. Confirmatory factor analysis led to re-specification of the model, eliminating three items. Construct validity was demonstrated on the basis of correlationanalysis. Instrument reliability reached satisfactory levels. Weaknesses and strengths of theresearch are pointed out. A new agenda for future research is suggested. / El objetivo del estudio consistió en validar la Escala de Cinismo Organizacional de Brandespara su empleo con trabajadores argentinos. Se analizó la equivalencia funcional del instrumento. Se administró una versión adaptada a una muestra de 396 empleados, quienes respondieron también una selección de instrumentos para explorar confianza en la organización, compromiso organizacional y satisfacción laboral. El análisis factorial exploratorio mostró una estructura de tres factores que explicaron el 68.8% de la varianza total. El análisis factorial confirmatorio condujo a re-especificar el modelo, eliminando tres ítems. La validez de constructo quedó demostrada en base a los análisis de correlación efectuados. La confiabilidad del instrumento alcanzó niveles satisfactorios. Se puntualizan las debilidades y fortalezas de la investigación realizada. Se sugiere una agenda para futuros estudios.

Esta pesquisa foi conduzida com o intuito de monitorar a confiabilidade e a validade do instrumento de classificação de pacientes proposto por PERROCA (1996), baseado nas necessidades individualizadas de cuidado de enfermagem. O proceso de validação compreendeu 2 etapas: o teste de confiabilidade entre avaliadores e o teste de validade de constructo. O instrumento de classificação de pacientes foi aplicado, na primeira etapa, pelas enfermeiras lotadas na Unidade de Terapia Intensiva Geral de um hospital de ensino, em 50 pacientes escolhidos aleatoriamente. Na etapa posterior, 141 pacientes foram avaliados por enfermeiras de Unidades de Internação e Unidades de Terapia Intensiva da mesma instituição. A avaliação do grau de concordância nos escores totais obtidos pelas enfermeiras foi verificado mediante a utilização do coeficiente de correlação de Pearson e do coeficiente Kappa. Para viabilizar a demonstração da validade de constructo do instrumento recorreu-se a múltiplas técnicas, tais como: Pearson (para estudar a associação entre os 13 indicadores); Análise de Componentes Principais e Análise de Variância (para verificar a possibilidade de redução do número de indicadores que compõem o instrumento) e Regressão Logística Ordinal (para determinar a capacidade preditiva do instrumento em captar os diferentes níveis de cuidado). A análise de confiabilidade revelou que as correlações são altas e positivas entre os escores dados pelas enfermeiras, apresentando um Kappa entre 0,68 e 0,90 com IC de 95%, indicando uma boa concordância. Esses métodos demonstraram a validade de construção do instrumento, apotando: 1 - todos os indicadores críticos como sendo indicadores do grau de complexidade do paciente; 2 - nítida diferenciação entre as categorias de cuidados, criando gradiente de complexidade; 3 - o indicador crítico Cuidado Corporal como o mais importante para captar a mudança na categoria de cuidado e o indicador Sinais Vitais como o de menor importância; 4 - alto valor preditico do instrumento, 99,8%; 5 - adequação do método de padronização do escore utilizado para delimitar categorias de cuidados. O estudo permitiu concluir que este instrumento apresenta evidências de confiabilidade e validade estando em condições de ser aplicado na prática gerencial do enfermeiro como norteador das necessidades de cuidado do paciente, bem como da carga de trabalho da equpe de enfermagem. / The purpose of this research was to check the reability and validity of PERROCA´s instrument model for classifying patients, based on the particular needs of nursing. The instrument process was divided into two parts: the reability assessment by the valuators and the assessment of the construct validity. At the first part, that patients´classification instrument was used in 50 patients randomly chosen by the nurses of the General Intensive Care Unit of a medical school. At the second part, 141 patients were evaluated by the nurses in the Intensive Care and Hospitalization Units of the same institution. Pearson´s quotient correlation and Kappa´s quotient were used to observe the concordance degree of the total scores by the nurses. Several techniques were used to show the construct validity of the instrument; such as: Pearson (to study the relationship among the 13 indicators); Principal Component Analysis and Analysis of Variance (to detrmine further any reduction of the instrument´s number of indicators) and Ordinal Logistic Regression (to obsereve the instrument´s predictive ability in drawing different levels of care). The reability analysis showed a high and positive correlation in relation to scores given by the nurses; that is, a Kappa between 0.68 and 0.90 with 95% of CI, showing a good cocordance. The construct validity of the instrument was showed by: 1- all critical indicators as being indicators of the patients´complexity degree; 2- a clear differentiation in the categories, producing a complexity rate; 3- Corporeal Hygiene, as a critical indicator, as the one most important to observe the change in the care category and the Vital Signs indicator as the one of the least importance; 4- high preditive value of the instrument, 99.8%; 5- adequacy of the method to standardize the used score for outling the care categories. This study showed that the instrument model has drawn evidence of reability and validity, therefore being useful for nursing management practice as a guide to the patients´s care needs as well as to the workload of the nursing team.

Os principais estudos sobre o ambiente familiar de pacientes com esquizofrenia são aqueles relacionados ao conceito de Emoção Expressa (EE), que trata da qualidade de interação social entre os membros da família. Existem poucos estudos que investigam o ambiente familiar do paciente com esquizofrenia no Brasil e nenhum instrumento de fácil aplicação para avaliar esse conceito. Trata-se de um estudo metodológico com o objetivo de adaptar e validar o Family Questionnaire (FQ) para língua portuguesa. O FQ contém 20 itens distribuídos em dois domínios, 10 itens para comentários críticos (CC) e 10 itens para super envolvimento emocional (SEE), com respostas de um (1) muito raramente a quatro (4) sempre, com um intervalo possível para o total da medida de cada domínio de 10 a 40, no qual maiores valores para os dois domínios refletem maior EE. O processo de adaptação cultural seguiu os passos preconizados pela literatura: tradução do FQ para língua portuguesa; obtenção do consenso das versões em português; avaliação pelo comitê de especialistas; Back-translation; obtenção do consenso das versões em alemão e comparação com versão original em alemão; avaliação semântica dos itens e pré-teste. Participaram do estudo 100 familiares de pacientes com esquizofrenia. Os dados foram coletados por meio de entrevistas, no período de janeiro a março de 2010, em quatro serviços secundários de saúde mental do interior do Estado de São Paulo. Para a análise de validade de constructo do FQ utilizou-se testes de correlação de Pearson entre as medidas dos domínios do FQ e de um constructo correlato (senso de coerência); análise do componente principal e teste para comparação de grupos distintos. A confiabilidade foi avaliada pela consistência interna (alfa de Cronbach) e a reprodutibilidade através do teste-reteste. O nível de significância adotado foi de 0,05. Os resultados apontaram que os participantes eram, predominantemente, do sexo feminino (78%), com idade média de 52,6 anos, com primeiro grau incompleto (52%), casados (57%), mães (41%) e com tempo médio de convivência com o paciente de 26,7 anos. Entre os pacientes, 66% eram do sexo masculino com idade média de 40,5 anos, a maioria com 1º grau incompleto (60%) e com tempo médio de doença igual há 15,8 anos. Os testes de hipóteses confirmaram a validade de constructo convergente entre as medidas do senso de coerência, e os domínios do FQ e, parcialmente, a diferença nas medidas dos domínios do FQ e as características dos familiares. A análise de componentes principais confirmou a mesma dimensionalidade da versão original na versão adaptada para o português. Em relação à confiabilidade, obtivemos valores adequados de de Cronbach para os domínios CC (0,861) e SEE (0,655) da versão adaptada do FQ. A reprodutibilidade avaliada pelo teste-reteste em 30 participantes apontou fortes correlações entre os itens dos domínios CC(r=0,689; p<0,001)e SEE (r=0,769; p<0,001) do FQ. Conclui-se que a versão adaptada para o português do FQ mostrou-se válida e confiável na amostra estudada. / The main studies about family environment of schizophrenic patients are those related to the concept of Expressed Emotion (EE), based on the quality of social interaction among family members. Few studies have investigated the family environment of schizophrenic patients in Brazil. There isn\'t an easy instrument to apply to evaluate this concept. This methodological study aimed to adapt and validate the Family Questionnaire (FQ) for the Portuguese language. The FQ contains 20 items distributed in two domains; 10 items for critical comments (CC) and 10 items for emotional overinvolvement (EOI). The items responses are obtained from a 4 point scale with extreme values varying from one (1) - very rarely - to four (4) - always. The possible interval for the total measure varies from 10 to 40, in which the higher values for both areas reflect higher EE. The process of cultural adaptation followed the steps recommended in the literature: translation of the FQ for Portuguese; obtaining consensus of the Portuguese versions; evaluation by expert committee, back-translation; obtaining consensus of versions in German and compared with the original version in German; semantic evaluation of the items and pre-test. Study participants were 100 relatives of schizophrenic patients. Data were collected through interviews in four secondary mental health services at São Paulo state during the period from January to March 2010. The validation of the FQ construct was evaluated through Pearson correlation tests between the measures of the instrument domain and a correlate construct (sense of coherence), principal component analysis and testing to compare different groups. Reliability was evaluated by internal consistency of its items (Cronbach\'s alpha) and reproducibility by test-retest. The level of significance adopted was of 0.05. The results showed that the majority of the participants were female (78%), average age of 52.6 years old, basic level of school (52%), married (57%), mothers (41%) and average time living with the patient of 26.7 years. Among the patients, 66% were male, average age of 40.5 years old, incomplete basic education (60%) and average time of illness of 15.8 years. Hypothesis tests confirmed the convergent construct validity between the measures of sense of coherence and the domains\' of FQ, and, in part, the difference in measurements of FQ domains\' and the characteristics of the families. The analysis of main components confirmed the same dimensionality from the original version in the version adapted for Portuguese. Regarding the reliability, we obtained an adequate Cronbach\'s alpha value for the fields CC (0.861) and EOI (0.655) of the FQ adapted version. Reproducibility evaluated by test-retest with 30 participants indicated strong correlations between items of the CC domains (r = 0.689, p <0.001) and EOI (r = 0.769, p <0.001) of the FQ. Therefore, we conclude the FQ Portuguese adapted version is valid and reliable for the sample studied.

Este estudo teve por finalidade a construção e validação de um instrumento para classificação de pacientes baseado nas necessidades individualizadas de cuidado de enfermagem. Para compor o instrumento foram considerados 13 indicadores críticos : Estado Mental e Nível de Consciência, Oxigenação, Sinais Vitais, Nutrição e Hidratação, Motilidade, Locomoção, Cuidado Corporal, Eliminações, Terapêutica, Educação à Saúde, Comportamento, Comunicação e Integridade Cutâneo-Mucosa. Cada um desses indicadores possui gradação de 1 a 5, apontando a intensidade crescente da complexidade assistencial. O paciente é classificado em todos os indicadores em um dos 5 níveis, na opção que melhor descreva a sua situação. Para validação do conteúdo foi aplicada a Técnica Delphi em 2 fases. Participaram como juízes 15 profissionais da área de enfermagem que atuam junto a instituições de assistência ou vinculados a Hospital Escola na cidade de São José do Rio Preto. Os resultados obtidos mostraram concordância dos juízes quanto a : manutenção dos 13 indicadores críticos no instrumento; pertinência e clareza do conteúdo dos indicadores críticos e a existência de nível de complexidade assistencial crescente / This study had in its aim the construction and the validation of a patient classification instrument which has been based on the patient’s individual necessities that require the nursing care. It was considered in the instrument 13 critical indicators : Mental State and Level of Conciousness, Breathing, Vital Signs, Nutrition and Hydration, Movement, Locomotion, Corporeal Hygiene, Eliminations, Therapy, Health Teaching, Behavior, Communication and Skin Integrety. Each one of these indicators has a 1 to 5 gradation denoting an increasing level in the nursing care complexity. The patient is classified in all the indicators in one of the five levels, in the option that better describes his/her situation. The content validation of the instrument was done by the Delphi Tecnique aplication through 2 rounds. A team of 15 nursing experts who attend pacients or teach in the Medical School in São José do Rio Preto were participants in this research. The obtained results have showed the experts’ agreement concerned to: the maintenance of the 13 critical indicators in the instrument; property and intelligibility of the critical indicator contents and the presence of an increasing level in the nursing care complexity

Objetivo. O principal objetivo deste estudo foi desenvolver um questionário de freqüência alimentar- QF AA para calcular o consumo de energia, carboidrato, proteína, lipídio total, lipídio insaturado, fibra, colesterol, retinol, vitamina C, cálcio e ferro de adolescentes freqüentadores de uma escola pública localizada na região oeste da cidade de São Paulo. A validade relativa do QF AA foi testada por meio da comparação entre os valores estimados pelo método de referência (média de 3 dias de R24h) e os calculados pelo QFAA. Metodologia. Depois de identificados os principais alimentos fontes dos nutrientes analisados, o QF AA foi testado em prova piloto. A versão final foi constituída por 76 itens alimentares, agrupados segundo o perfil de cada alimento. O estudo de validade foi realizado durante um período de 6 meses (junho a novembro de 1999) junto a 79 alunos voluntários de ambos os sexos, os quais responderam 4 R24h com intervalo de 45 dias e 1 QFAA ao final do estudo. Os dados foram calculados pelo Virtual Nutri e analisados utilizando-se coeficiente de correlação de Pearson, ajustados pela caloria e corrigidos pela variância intrapessoal. Resultados. Observou-se uma alta variabilidade no consumo da dieta dos adolescentes, mostrando razões de variância extremamente altas para o colesterol, retinol e vitamina C. Os coeficientes de correlação de Pearson, depois do ajuste e da correção pela variância, variaram entre - O,1O a 0,72 para os indivíduos de sexo feminino e de O,16 a 0,91 para os de sexo masculino. A média do coeficiente de correlação para todo o grupo foi de 0,52. Conclusões. Nossos resultados indicam que o QFAA mostrou aceitável desempenho para classificar os indivíduos segundo seu consumo habitual para a maioria dos nutrientes, com exceção do retinol e do ferro. / Purpose. The main purpose of this study was to develop a Food-Frequency Questionnaire - FFQ to calculate the consumption of energy, carbohydrate, protein, total lipid, unsaturated lipid, flber, cholesterol, retinol, vitamin C, calcium and iron by adolescents attending a public school located in the western region of the city of São Paulo. The relative validity of the FFQ was tested in a comparison between values estimated by the reference method (average, three days of R24h) and those calculated by the FFQ. Methodology. After identifying the major food sources of the nutrients analyzed, the FFQ was tested in a pilot test. The final version was comprised by seventy-six food items, grouped according to each of their proflles. The validity study was performed over a six-month period (]une through November, 1999) on seventy-nine volunteer students of both sexes, which responded four R24h with a 45-day interval and one FFQ at the end of the study. Data was computed by the Virtual Nutri and analyzed by using Pearson\'s correlation coefflcient, energy - adjustment and corrected for within-subject variance. Results. High variability was observed in adolescent diet consumption, showing extremely high variance ratios for cholesterol, retinol and vitamin C. Pearson\'s correlation coefficients, after adjustment and correction for variance, range between - O.1O to O.72 for females and from O.16 to O.91 for males. The average correlation coefficient for the entire group was 0.52. Conclusions. Our results indicate that the FFQ has shown an acceptable performance to classify individuais according to their habitual consumption for the majority of nutrients, with the exception of retinol and iron.

Abstract This interdisciplinary study describes in the first part the frequency with which various statistical research designs and methods are reported in psychiatric journals, and investigates how the use of these methods affect the visibility of the article in the form of received citations. In the second part focus is specifically on factor analysis, and the study presents two applications of this method. Original research articles (N = 448) from four general psychiatric journals in 1996 were reviewed. The journals were the American Journal of Psychiatry, the Archives of General Psychiatry, the British Journal of Psychiatry and the Nordic Journal of Psychiatry. There were differences in the utilisation of statistical procedures among the journals. The use of statistical methods was not strongly associated with the further utilisation of an article. However, extended description of statistical procedures had a positive effect on the received citations. Factor analysis is a statistical method based on correlations of the variables, which is often used when validity and structure of psychiatric instruments are studied. Exploratory factor analysis is designed to explore underlying latent factors, and in confirmatory factor analysis the aim is to verify the factor structure based on earlier findings in other data sets. Using data from the 31-year follow-up of the Northern Finland 1966 Birth Cohort Study this study aimed to demonstrate the validity and factor structure of scales measuring temperament (Tridimensional Personality Questionnaire, TPQ, and Temperament and Character Inventory, TCI) and alexithymia (20-item Toronto Alexithymia Scale, TAS-20). The results of exploratory factor analysis indicated good performance of the TCI and TPQ, though the results suggested that some developmental work is still needed. Of the two scales, the TCI worked psychometrically better than the TPQ. A confirmatory factor analysis showed that the three-factor model of TAS-20 was in agreement with the Finnish version of the scale. To conclude, future authors of psychiatric journals might apply these results in designing their research to present intelligible and compact analysis combined with a high quality presentation technique. Results of the factor analyses showed that the TPQ, TCI and TAS-20 can be used also in their Finnish versions. / Tiivistelmä Tämä poikkitieteellinen tutkimus kuvaa erilaisten tilastotieteellisten menetelmien yleisyyttä ja merkitystä psykiatriassa. Tutkimuksen ensimmäisessä osassa tutkitaan erilaisten tilastomenetelmien ja tutkimusasetelmien osuutta psykiatrisissa artikkeleissa ja lisäksi käytettyjen menetelmien vaikutusta artikkelien saamien viittausten lukumäärään. Tutkimuksen toisessa osassa keskitytään faktorianalyysiin ja esitetään kaksi siihen liittyvää sovellusta. Aineiston muodostavat alkuperäistuloksia esittelevät artikkelit (N = 448) neljästä eri psykiatrian tieteellisestä yleislehdestä vuodelta 1996. Kyseiset lehdet ovat American Journal of Psychiatry, Archives of General Psychiatry, British Journal of Psychiatry ja Nordic Journal of Psychiatry. Lehdet erosivat toisistaan tilastotieteellisten menetelmien käytössä ja tulosten esittämisessä. Tilastotieteellisten menetelmien käytöllä ei ollut suurta vaikutusta artikkelien saamien viittausten lukumäärään, mutta laajalla menetelmien kuvauksella oli positiivinen vaikutus viittausten lukumäärään. Faktorianalyysi on tilastotieteellinen tutkimusmenetelmä, jota käytetään tutkittaessa millaisista osatekijöistä erilaiset monimutkaiset ilmiöt koostuvat. Erityisesti tutkittaessa psykiatristen mittareiden validiteettia ja rakennetta faktorianalyysi on osoittautunut hyödylliseksi. Eksploratiivisessa faktorianalyysissa tarkoituksena on etsiä taustalla olevia piileviä muuttujia ja konfirmatorisessa faktorianalyysissa tarkoitus on vahvistaa aiemmissa tutkimuksissa todettu mittarin faktorirakenne. Tässä tutkimuksessa hyödynnetään aineistoa Pohjois-Suomen vuoden 1966 syntymäkohortin 31 vuoden seurannasta. Aineiston avulla tutkitaan temperamenttia (Tridimensional Personality Questionnaire, TPQ, ja Temperament and Character Inventory, TCI) ja aleksitymiaa (20-item Toronto Alexithymia Scale, TAS-20) tutkivien mittareiden suomenkielisten käännöksien validiteettia ja faktorirakennetta. Eksploratiivisen faktorianalyysin tulokset kertoivat, että TPQ ja TCI toimivat hyvin myös suomenkielellä. Kuitenkin mittareissa on vielä kehittämisen varaa. TCI:n psykometriset ominaisuudet olivat paremmat kuin TPQ:n. Aleksitymiamittarin TAS-20 konfirmatorinen faktorianalyysi osoitti että aiemmin julkaistu kolmen faktorin malli toimi hyvin myös suomalaisella versiolla. Psykiatristen artikkelien kirjoittajat voivat hyödyntää tämän tutkimuksen tuloksia suunnitellessaan psykiatrista tutkimusta suuntaan, jossa selkeä ja tiivis tulosten analysointitapa ja korkealaatuinen tulosten esitystapa korostuu. Faktorianalyysi soveltuu hyvin mittarin validiteetin tutkimiseen. Tutkimus osoitti TPQ-, TCI- ja TAS-20-mittareiden suomenkielisten versioiden validiteetin.

Introduction: The Antimicrobial Self-assessment Toolkit for Acute NHS Trusts (ASAT) was developed by a pharmacist reference group of an Advisory Non-Departmental Public Body on Antimicrobial Resistance and Healthcare Associated Infections (ARHAI). It was developed in conjunction with the Department of Health. The primary purpose of the ASAT is to identify and to measure the methods of implementation of antimicrobial stewardship programmes in acute NHS trusts. The face validity was previously tested by ARHAI. The overall aims of this programme of work were to investigate the validity of the ASAT and to make iterative changes to improve its validity. Ethical approval was not required for this PhD project because it was categorised as service evaluation by the LREC. Also, ethical approval from the University of Manchester Research Ethics Committee was deemed unnecessary at the time of the PhD project due to the nature of the data collected. Methods: A mixed methodology approach utilising a sequential exploratory strategy was used to investigate the validity of the ASAT. This PhD project was composed of four sequential studies which resulted in iterative changes to the ASAT, that is, from ASAT v15a to ASAT v18. In Study 1, cognitive interviews were conducted with eight antimicrobial pharmacists in order to investigate the content validity of ASAT v15a. In Study 2, both cognitive interviews and semi-structured interviews were conducted with 10 clinical microbiologists in order to investigate the content validity of ASAT v16. In Study 3, Rasch modelling and analyses using the Partial Credit Model (PCM) were conducted on the responses to ASAT v17 from 33 NHS trusts across England. In Study 4, simple OLS regression analyses were conducted using the NHS trust ‘ability’ estimates or calibrations and Clostridium difficile (CDI) rates of participating NHS trusts in order to investigate model fit and the predictive validity of the ASAT. Results: The cognitive interviews conducted in study 1 indicated that AMPs encountered cognitive difficulties along the cognitive processing pathway in response to ASAT v15a. These difficulties included comprehension in 27 (32.5%) questions and response generation/formatting in 13 (15.7%) questions. Also respondents indicated that the role of clinical microbiologists in ASPs was underrepresented in ASAT v15a. The interviews conducted in Study 2 were confirmatory in nature as they reflected the findings of Study 1. For example terms such as ‘formulary’ and ‘policy’ were misinterpreted by respondents. Rasch modelling and analysis showed that there were items within ASAT v17 which were underfitting and overfitting the Partial Credit Model. Item fit was investigated after removal of these items which resulted in improved fit for domains 2 and 5. ASAT v18 was developed after these analyses and was included items that were productive for measurement. On examination of the OLS regression analyses conducted in Study 4, it was seen that there was poor model fit and very limited predictive validity of the model. Conclusion: The iterative methodology utilised to investigate the validity and subsequently improve the ASAT was effective in establishing content and construct validity. However, the predictive validity of the ASAT was limited. This may be due to the outcome variable chosen for the OLS regression modelling. A more sensitive outcome measure such as compliance to treatment or prophylaxis guidelines may have been more effective at establishing predictive validity. The findings of this programme of work highlighted that there is further work required to validate the ASAT such as the determination of the appropriate weights and scores for ASAT domains and also the determination of the appropriate outcomes measures to determine the efficacy of ASPs. It is recommended that further validity testing should be conducted before a further iteration of the ASAT is used as a set of quality standards or as a hospital benchmarking tool

Psychometric instruments are used internationally in organisations to assist companies when making recruitment or development decisions. But the differences that exist on similar psychometric instruments between international groups of people with similar qualifications necessitate the evaluation of the construct validity of these instruments. The objective of this study was to determine the construct validity of the locus of control inventory developed by Udai Pareek in 1998. The instrument consists of 30 items and it was completed by 155 pre-and postgraduate students at the University of Pretoria. Oblimin rotation was used to assist with the interpretation of the factors. In the pattern and structure matrix the highest loadings were highlighted which meant that these items measured the construct that they were supposed to measure. Items were deleted where there was no clear indication as to what the item was measuring. Item analysis was done on each of the constructs identified to further investigate the appropriateness of each item and even more items were deleted, which also had an impact on the Cronbach’s alpha value. The chi-square as well as other fit indices was used to determine the model fit. Confirmatory factor analysis was conducted to establish how well the model fitted the data. It was clear that the two-factor model fitted the data considerably better than the three-factor model because of a high correlation between two of the factors which indicated that they might be measuring the same construct. The results indicate that the locus of control instrument is not suitable for use in the South African context and should therefore be modified until a satisfactory model fit is found. AFRIKAANS: Psigometrise instrumente word wereldwyd benut om maatsappye te help met werwing en ontwikkelings besluite. Maar die verskille wat ontstaan wanneer internasionale groepe met soortgelyke kwalifikasies teen mekaar opgeweeg word vereis dat die konstruk geldigheid van die instrumente bepaal word. Die doel van die studie was om die konstruk geldigheid van die lokus van kontrole vraelys, wat deur Udai Pareek ontwikkel is, te ondersoek.Die instrument bestaan uit 30 items en dit was voltooi deur 155 voor-en nagraadse studente van die Universiteit van Pretoria. Oblimin rotasie was gebruik om die interpretasie van die faktore te vergemaklik. Die items wat werklik meet wat dit ontwikkel is om te meet is ingekleur sodat dit maklik geidentifiseer kan word.Items is slegs uit die vraelys gehaal waar daar geen duidelike indikasie was wat die item meet nie. Item analise was op elke konstruk toegepas, en die chi-square en ander passings indekse was gebruik om te bepaal hoe die model op die data pas. Dit was duidelik dat die twee-faktor model die data baie beter pas as die voorgestelde drie-faktor model. Volgens die drie faktor model is daar twee van die voorgestelde drie faktore wat moontlik die selfde konstruk kan meet. Die resultate dui aan dat die lokus van kontrole vraelys nie geskik is vir gebruik in Suid–Afrika nie, en sal daarom aangepas moet word totdat „n aanvaarbare model gevind is. Copyright 2009, University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. Please cite as follows: Engelbrecht, C 2009, Determining the construct validity of Udai Pareek's locus of control inventory , MCom dissertation, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-02092012-100414 / > C12/4/75/gm / Dissertation (MCom)--University of Pretoria, 2012. / Human Resource Management / unrestricted