Bayesian inference for a skew-normal IRT model under the centred parameterization

Item response theory (IRT) comprises a set of statistical models which are useful in many fields, especially when there is interest in studying latent variables. These latent variables are directly considered in the Item Response Models (IRM) and they are usually called latent traits. A usual assumption for parameter estimation of the IRM, considering one group of examinees, is to assume that the latent traits are random variables which follow a standard normal distribution. However, many works suggest that this assumption does not apply in many cases. Furthermore, when this assumption does not hold, the parameter estimates tend to be biased and misleading inference can be obtained. Therefore, it is important to model the distribution of the latent traits properly. In this paper we present an alternative latent traits modeling based on the so-called skew-normal distribution; see Genton (2004). We used the centred parameterization, which was proposed by Azzalini (1985). This approach ensures the model identifiability as pointed out by Azevedo et al. (2009b). Also, a Metropolis Hastings within Gibbs sampling (MHWGS) algorithm was built for parameter estimation by using an augmented data approach. A simulation study was performed in order to assess the parameter recovery in the proposed model and the estimation method, and the effect of the asymmetry level of the latent traits distribution on the parameter estimation. Also, a comparison of our approach with other estimation methods (which consider the assumption of symmetric normality for the latent traits distribution) was considered. The results indicated that our proposed algorithm recovers properly all parameters. Specifically, the greater the asymmetry level, the better the performance of our approach compared with other approaches, mainly in the presence of small sample sizes (number of examinees). Furthermore, we analyzed a real data set which presents indication of asymmetry concerning the latent traits distribution. The results obtained by using our approach confirmed the presence of strong negative asymmetry of the latent traits distribution. (C) 2010 Elsevier B.V. All rights reserved.

Bayesian Estimation of the Logistic Positive Exponent IRT Model

A Bayesian inference approach using Markov Chain Monte Carlo (MCMC) is developed for the logistic positive exponent (LPE) model proposed by Samejima and for a new skewed Logistic Item Response Theory (IRT) model, named Reflection LPE model. Both models lead to asymmetric item characteristic curves (ICC) and can be appropriate because a symmetric ICC treats both correct and incorrect answers symmetrically, which results in a logical contradiction in ordering examinees on the ability scale. A data set corresponding to a mathematical test applied in Peruvian public schools is analyzed, where comparisons with other parametric IRT models also are conducted. Several model comparison criteria are discussed and implemented. The main conclusion is that the LPE and RLPE IRT models are easy to implement and seem to provide the best fit to the data set considered.

Essays on consumer brand engagement

Apesar do crescente interesse no conceito de engajamento da marca ainda existe discordância quanto aos seus conceitos fundamentais. Esta tese de doutorado explora a natureza da construção engajamento da marca do consumidor (EMC). No primeiro artigo, EMC é avaliada no âmbito da Teoria da Expectância para explicar e esclarecer como a antecipação de possíveis resultados de se envolver com uma marca, sendo tais resultados classificados como “primeiro nível” (resultante do esforço pessoal alocado para interagir com uma marca) e “segundo nível” (ou nível final, representando a consequência dos resultados de primeiro nível) e uma nova definição de EMC é formulada. Um arcabouço teórico abrangente é proposto para engajamento da marca, usando o Teoria Organizacional de Marketing para Expansão de Fronteiras (TOMEF) como referência para os pontos de contato entre o consumidor e a marca. A partir dos fundamentos teóricos das dimensões cognitivas, emocionais e comportamentais do EMC, quinze proposições teóricas são desenvolvidas para incorporar uma perspectiva multilateral às doutrinas teóricas do construto. No segundo artigo, quatro estudos são usados para desenvolver uma escala de engajamento da marca do consumidor. O Estudo 1 (n = 11) utiliza revisão da literatura e entrevistas em profundidade com os consumidores para gerar os itens da escala. No Estudo 2, oito especialistas avaliam 144 itens quanto a validade de face e validade de conteúdo. No Estudo 3 dados coletados com alunos de graduação (n = 172) é submetida à análise fatorial exploratória (AFE) e confirmatória (AFC) para redução adicional de itens. Trezentos e oitenta e nove respostas de um painel de consumidores são usados no Estudo 4 para avaliar o ajuste do modelo, usando a análise fatorial confirmatória (AFC) e Modelagem por Equações Estruturais (MEE). A escala proposta possui excelentes níveis de validade e confiabilidade. Finalmente, no terceiro papel, uma escala de engajamento do consumidor de Vivek et al. (2014) é replicada (n = 598) junto à consumidores em uma feira automotiva, para estender o debate sobre formas de medição do constructo usando a perspectiva da Teoria de Resposta ao Item (TRI). Embora o modelo desenvolvido com base na teoria clássica de teste (TCT) usando AFC, um modelo de resposta gradual (MRG) identifica cinco itens que têm baixos níveis de poder discriminante e com baixos níveis de informação. A abordagem usando TRI indica um possível caminho para melhorias metodológicas futuras para as escalas desenvolvidas na área de marketing em geral, e para a escala engajamento do consumidor, em particular.

Novas abordagens na avaliação e construção de rankings

Esta tese apresenta algumas abordagens ainda não exploradas na avaliação e construção de rankings, exclusivamente daqueles baseados em indicadores compostos. Para isso, três artigos foram desenvolvidos com o intuito de evoluir com uma literatura genericamente aplicável, ou seja, não restrita a contextos de rankings específicos. No primeiro desses artigos, composto por três estudos, mostrou-se que as informações percebidas pelos usuários através dos rankings nem sempre são fornecidas por eles. No segundo, o qual pode ser entendido como uma extensão do primeiro, propôs-se a criação de uma métrica – intitulada COMP – destinada a mensurar o grau de compatibilidade entre as informações percebidas pelos usuários e aquelas fornecidas pelos rankings. No terceiro artigo, independente dos dois primeiros, explorou-se o potencial da Teoria de Resposta ao Item (TRI) enquanto metodologia para a avaliação e construção de rankings. Para isso, dois estudos, o primeiro deles focado no Failed States Index (FSI) e o segundo no Index of Economic Freedom (IEF) foram desenvolvidos para mostrar as potencialidades da metodologia proposta.

Conhecimento de escolares sobre leishmaniose visceral

Pós-graduação em Ciência Animal - FMVA

Planejamento de um ambiente virtual de aprendizagem baseado em interfaces dinâmicas e uma aplicação ao estudo de potência elétrica

Pós-graduação em Engenharia Elétrica - FEIS

Competitividade inter-regional de sistemas de produção de bovinocultura de corte

O trabalho apresenta uma metodologia de avaliação de sistemas de produção através da mensuração da competitividade interna na bovinocultura de corte. Durante o primeiro trimestre de 2010, foram aplicados 65 questionários com pecuaristas, sendo 36 entrevistas na Região Sul (Estado do Rio Grande do Sul) e 29 na Região Norte (Estados do Pará e Rondônia). Foram definidos os principais direcionadores que afetam a competitividade interna - tecnologia, gestão, relações de mercado e ambiente institucional, sendo atribuído um peso específico para cada direcionador, a fim de obter o índice de competitividade. Os resultados foram analisados estatisticamente pela teoria de resposta ao item e pela análise de correspondência (ANACOR) com o software SPSS®. A Região Sul apresentou uma maior competitividade que a Região Norte. Independente da região, os fatores críticos de competitividade foram: integração lavoura-pecuária, planejamento estratégico, cálculo de indicadores financeiros, formação de preços, acesso a inovações tecnológicas e organização dos produtores.

Criação de um banco de itens para avaliar o conhecimento do câncer de cabeça e pescoço

Pós-graduação em Odontologia Preventiva e Social - FOA

Construção de uma escala para avaliação da pro ciência em conteúdos matemáticos básicos

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Atividades de educação em saúde sobre leishmaniose visceral para escolares

This study aimed at evaluating the knowledge on visceral leishmaniasis gained after the application of an educative project for the 6th and 7th grade students from three public schools of Birigui, SP, Brazil. A questionnaire before (Phase I) and after (Phase II) activities that comprehended one conference by a health agent, a comic contest and one crossword about VL was used to measure scholar’s knowledge. We interviewed 711 students in Phase I and 693 in Phase II. A criterion of VL knowledge was adopted as “Good”, “Medium” and “Bad” when, out of 10 questions analyzed by Item Response Theory, 10 to 8, 7 to 4, and 3 to 0 were right, respectively. We observed a statistically significant increase in the students’ knowledge level after the educational project, since the number of students with “Good” concept changed from 35.7% (Phase I) to 59.7% (Phase II). The educational activities carried out led to gains in knowledge among students suggesting that continuing education can bring good results to public health.

Parameter recovery for a skew-normal IRT model under a Bayesian approach: hierarchical framework, prior and kernel sensitivity and sample size

Item response theory (IRT) comprises a set of statistical models which are useful in many fields, especially when there is an interest in studying latent variables (or latent traits). Usually such latent traits are assumed to be random variables and a convenient distribution is assigned to them. A very common choice for such a distribution has been the standard normal. Recently, Azevedo et al. [Bayesian inference for a skew-normal IRT model under the centred parameterization, Comput. Stat. Data Anal. 55 (2011), pp. 353-365] proposed a skew-normal distribution under the centred parameterization (SNCP) as had been studied in [R. B. Arellano-Valle and A. Azzalini, The centred parametrization for the multivariate skew-normal distribution, J. Multivariate Anal. 99(7) (2008), pp. 1362-1382], to model the latent trait distribution. This approach allows one to represent any asymmetric behaviour concerning the latent trait distribution. Also, they developed a Metropolis-Hastings within the Gibbs sampling (MHWGS) algorithm based on the density of the SNCP. They showed that the algorithm recovers all parameters properly. Their results indicated that, in the presence of asymmetry, the proposed model and the estimation algorithm perform better than the usual model and estimation methods. Our main goal in this paper is to propose another type of MHWGS algorithm based on a stochastic representation (hierarchical structure) of the SNCP studied in [N. Henze, A probabilistic representation of the skew-normal distribution, Scand. J. Statist. 13 (1986), pp. 271-275]. Our algorithm has only one Metropolis-Hastings step, in opposition to the algorithm developed by Azevedo et al., which has two such steps. This not only makes the implementation easier but also reduces the number of proposal densities to be used, which can be a problem in the implementation of MHWGS algorithms, as can be seen in [R.J. Patz and B.W. Junker, A straightforward approach to Markov Chain Monte Carlo methods for item response models, J. Educ. Behav. Stat. 24(2) (1999), pp. 146-178; R. J. Patz and B. W. Junker, The applications and extensions of MCMC in IRT: Multiple item types, missing data, and rated responses, J. Educ. Behav. Stat. 24(4) (1999), pp. 342-366; A. Gelman, G.O. Roberts, and W.R. Gilks, Efficient Metropolis jumping rules, Bayesian Stat. 5 (1996), pp. 599-607]. Moreover, we consider a modified beta prior (which generalizes the one considered in [3]) and a Jeffreys prior for the asymmetry parameter. Furthermore, we study the sensitivity of such priors as well as the use of different kernel densities for this parameter. Finally, we assess the impact of the number of examinees, number of items and the asymmetry level on the parameter recovery. Results of the simulation study indicated that our approach performed equally as well as that in [3], in terms of parameter recovery, mainly using the Jeffreys prior. Also, they indicated that the asymmetry level has the highest impact on parameter recovery, even though it is relatively small. A real data analysis is considered jointly with the development of model fitting assessment tools. The results are compared with the ones obtained by Azevedo et al. The results indicate that using the hierarchical approach allows us to implement MCMC algorithms more easily, it facilitates diagnosis of the convergence and also it can be very useful to fit more complex skew IRT models.

Psychometric properties of the Beck Depression Inventory-II: a comprehensive review

OBJECTIVE: To review the psychometric properties of the Beck Depression Inventory-II (BDI-II) as a self-report measure of depression in a variety of settings and populations. METHODS: Relevant studies of the BDI-II were retrieved through a search of electronic databases, a hand search, and contact with authors. Retained studies (k = 118) were allocated into three groups: non-clinical, psychiatric/institutionalized, and medical samples. RESULTS: The internal consistency was described as around 0.9 and the retest reliability ranged from 0.73 to 0.96. The correlation between BDI-II and the Beck Depression Inventory (BDI-I) was high and substantial overlap with measures of depression and anxiety was reported. The criterion-based validity showed good sensitivity and specificity for detecting depression in comparison to the adopted gold standard. However, the cutoff score to screen for depression varied according to the type of sample. Factor analysis showed a robust dimension of general depression composed by two constructs: cognitive-affective and somatic-vegetative. CONCLUSIONS: The BDI-II is a relevant psychometric instrument, showing high reliability, capacity to discriminate between depressed and non-depressed subjects, and improved concurrent, content, and structural validity. Based on available psychometric evidence, the BDI-II can be viewed as a cost-effective questionnaire for measuring the severity of depression, with broad applicability for research and clinical practice worldwide.

L'impiego del modello di Rash per la valutazione della soddisfazione del paziente nell'ambito delle strutture sanitarie di ricovero

Hospitals and health service providers are use to collect data about patient’s opinion to improve patient health status and communication with them and to upgrade the management and the organization of the health service provided. A lot of survey are carry out for this purpose and several questionnaire are built to measure patient satisfaction. In particular patient satisfaction is a way to describe and assess the level of hospital service from the patient’s point of view. It is a cognitive and an emotional response to the hospital experience. Methodologically patient satisfaction is defined as a multidimensional latent variable. To assess patient satisfaction Item Response Theory has greater advantages compared to Classical Test Theory. Rasch model is a one-parameter model which belongs to Item Response Theory. Rasch model yield objective measure of the construct that are independent of the set of people interviewed and of set of items used. Rasch estimates are continuous and can be useful to “calibrate” the scale of the latent trait. This research attempt to investigate the questionnaire currently adopted to measure patient satisfaction in an Italian hospital, completed by a large sample of 3390 patients. We verify the multidimensional nature of the variable, the properties of the instrument and the level of satisfaction in the hospital. Successively we used Rasch estimates to describe the most satisfied and the less satisfied patients.

A bayesian model for mixed responses and skew laten variable to measure HRQoL in children

The aim of the thesi is to formulate a suitable Item Response Theory (IRT) based model to measure HRQoL (as latent variable) using a mixed responses questionnaire and relaxing the hypothesis of normal distributed latent variable. The new model is a combination of two models already presented in literature, that is, a latent trait model for mixed responses and an IRT model for Skew Normal latent variable. It is developed in a Bayesian framework, a Markov chain Monte Carlo procedure is used to generate samples of the posterior distribution of the parameters of interest. The proposed model is test on a questionnaire composed by 5 discrete items and one continuous to measure HRQoL in children, the EQ-5D-Y questionnaire. A large sample of children collected in the schools was used. In comparison with a model for only discrete responses and a model for mixed responses and normal latent variable, the new model has better performances, in term of deviance information criterion (DIC), chain convergences times and precision of the estimates.

A quality of life subscale for terminally ill cancer patients

A subscale was developed to assess the quality of life of cancer patients with a life expectancy of six months or less. Phase I of this study identified the major concerns of 74 terminally ill cancer patients (19 with breast cancer, 19 with lung cancer, 18 with colorectal cancer, 9 with renal cell cancer, 9 with prostate cancer), 39 family caregivers, and 20 health care professionals. Patients interviewed were being treated at the University of Texas M. D. Anderson Cancer Center or at the Hospice at the Texas Medical Center in Houston. In Phase II, 120 patients (30 with breast cancer, 30 with lung cancer, 30 with colorectal cancer, 15 with prostate cancer, and 15 with renal cell cancer) rated the importance of these concerns for quality of life. Items retained for the subscale were rated as "extremely important" or "very important" by at least 60% of the sample and were reported as being applicable by at least two-thirds of the sample. The 61 concerns that were identified were formatted as a questionnaire for Phase III. In Phase III, 356 patients (89 with breast cancer, 88 with lung cancer, 88 with colorectal cancer, 44 with prostate cancer, and 47 with renal cell cancer) were interviewed to determine the subscale's reliability and sensitivity to change in clinical status. Both factor analysis and item response theory supported the inclusion of the same 35 items for the subscale. Internal consistency reliability was moderate to high for the subscale's domains: spiritual (0.87), existential (0.76), medical care (0.68), symptoms (0.67), social/family (0.66), and emotional (0.61). Test-retest correlation coefficients also were high for the domains: social/family (0.86), emotional (0.83), medical care (0.83), spiritual (0.75), existential (0.75), and symptoms (0.81).^ In addition, concurrent validity was supported by the high correlation between the subscale's symptom domain and symptom items from the European Organization for Research and Treatment of Cancer (EORTC) scale (r = 0.74). Patients' functional status was assessed with the Eastern Cooperative Oncology Group (ECOG) Performance status rating. When ECOG categories were compared to subscale domains, patients who scored lower in functional status had lower scores in the spiritual, existential, social/family, and emotional domains. Patients who scored lower in physical well-being had higher scores in the symptom domain. Patient scores in the medical care domain were similar for each ECOG category. The results of this study support the subscale's use in assessing quality of life and the outcomes of palliative treatment for cancer patients in their last six months of life. ^