P043. Aplicaciones de la Teoría de Respuesta al Ítem a la Evaluación del Razonamiento y de Actitudes y Estudios de Simulación para el Análisis del Funcionamiento Diferencial del Ítem.
Director/a: Attorresi, Horacio Félix
e-mail: horacioattorresi@fibertel.com.ar
Codirector: Aguerri, María Ester
e-mail: maguerri@psi.uba.ar
Codirector: Galibert, María Silivia
e-mail: galibert@psi.uba.ar
Unidad Ejecutora: FACULTAD DE PSICOLOGÍA
Área Temática: PSICOLOGIA
Rama: Psicología Educacional
Especialidad: Psicometría - Evaluación
Resumen en español:
Este proyecto tiene como finalidad el diseño y la construcción de instrumentos para la evaluación de aptitudes tales como el Razonamiento Deductivo y el Razonamiento Verbal y para la evaluación de las actitudes Voluntad de Trabajo y Altruismo. Se definirán los respectivos constructos y se emplearán diversos modelos de
Resumen en inglés:
APPLICATIONS OF THE ITEM RESPONSE THEORY TO THE ASSESSMENT OF REASONING AND ATTITUDES AND SIMULATION STUDIES FOR THE ANALYSIS OF DIFFERENTIAL ITEM FUNCTIONING.
The aim of this project is to design and construct instruments for measuring aptitudes such as Deductive Reasoning and Verbal Reasoning, and for the assessment of the attitudes Will-to-Work and Altruism. Their respective constructs will be defined and a variety of Item Response Models will be applied, which are: the Rasch one, the Fischer’s LLTM, the Two- and Three-Parameter Logistic Models and the Graded Response and Partial Credit Models. An item bank of verbal analogies will be built in order to be implemented in adaptive computerized tests. The identification of biases and the effect of sample size in the Probabilistic Reasoning will be studied. The Type I and II error rates of some statistical tests for detecting Uniform and Nonuniform Differential Item Functioning (DIF) will be simulated under different situations and will be compared. Those statistical tests are the Breslow-Day Tests, both the Global and the Trend of Odds Ratio Heterogeneity ones, and the Mantel-Haenszel Procedures. The results will be applied to detect DIF and to analyze the possible bias of the items that will be constructed.