Article

La evaluación de las consecuencias del uso de los tests en la teoría de la validez

Psicothema, ISSN 0214-9915, Vol. 18, Nº. 2, 2006, pags. 307-312 01/2006;
Source: OAI

ABSTRACT Evaluation of consequences of test use invalidity theory. There is little doubt about the importance of validity during the compilation and evaluation of tests. Nevertheless, intense debate has arisen with regard to incorporating the consequences of test use as a further source of evidence in the most recent edition of the AERA, APA, NCME Standards. After reviewing the historical antecedents of the issue and the main lines of approach of both defenders and critics, this paper sets out the arguments which may be used in answer to the question: 'At what point should the analysis of the consequences of test use become part of validation?' The response of the AERA, APA, NCME Standards has clear simila- rities with the perspective adopted towards the consideration of bias in tests as a problem of validity. Finally, the paper describes how consequence validation may be subject to tension arising from the evolution of notions of justice in test use.

0 Bookmarks
 · 
94 Views
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: In this article, it is assumed that gender violence, or violence against women, has mainly a sociocultural basis. A scale (Adolescent Sexism Detection; ASD) to detect sexism in adolescents was developed and its psychometric properties were analysed. 245 adolescents between 14 and 17 years of age participated in the study. As a result of the factor analyses carried out, the hostile and benevolent dimensions of sexism were clearly differentiated. Convergent validity of the scale was confirmed by its high correlations with the Ambivalent Sexism Inventory (ASI). Participants scored higher in benevolent than in hostile sexism, an effect even stronger in female participants. These findings suggest that the ASD scale is a valid instrument to measure sexism among adolescents. The main findings of this study are compared to those obtained in other studies using the ASI. The potential of this scale to appropriately detect sexism among adolescents is discussed.
    Psicothema 09/2007; 19(3):522-8. · 0.96 Impact Factor
  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: Resumen. El presente trabajo pretende generar evidencias empíricas en torno a la validez de la prueba de "Diagnóstico de conocimientos y destrezas en matemática del estudiante al ingresar a la universidad", de la Escuela de Matemática de la Uni-versidad de Costa Rica, desde la aplicación del modelo de Rasch. La muestra corresponde a 2624 examinados del 2008. Los objetivos del estudio se dirigieron primeramente a establecer evidencias de validez y confiabilidad para el instrumento. Por medio de análisis de factores exploratorio se verificó la unidimensionalidad de la escala y con el modelo de Rasch se generaron evidencias para concluir un grado aceptable de confiabilidad. Con la participación de 5 jueces expertos se establecieron niveles sustantivos de desempeño, clasificando los ítems según dificultad, y según procesos y contenidos necesarios para su resolución. Para validar las valoraciones de los jueces se contrastaron sus clasificaciones de dificultad con las estimaciones obtenidas al aplicar el modelo de Rasch, y por medio de un análisis de concordancia con la medida Kappa de Cohen se logró determinar el grupo de los 3 jueces que se acercaban más a las estimaciones de Rasch y cuyas valoraciones fueron consideradas para establecer los niveles de desempeño. Palabras clave: Pruebas estandarizadas, Matemática, validez,confiabilidad, modelos de Rasch, juicio experto, niveles de desem-peño. Abstract. The study intended, by means of the Rasch model, to provide empirical evidences regarding the validity of the test called "Diagnostic of knowledge and skills in Mathematics of the student entering the University", developed by the School of Mathematics at the University of Costa Rica. The sample consisted of 2624 examinees in the year 2008. The research objectives first addressed the issue of establishing validity and reliability evidences for the instrument. Using exploratory factor analysis the unidimensionality of the scale was confirmed, and employing the Rasch model evidence was generated to conclude an acceptable degree of reliability. With the participation of 5 expert judges substantive levels of performance were established, classifying the items according to difficulty, and according to necessary processes and contents for their solution. To validate the judges' assessments, their difficulty classifications were contrasted with the difficulty estimations from the Rasch model, and, making use of a concordance analysis with Cohen's Kappa the group of the 3 judges that were closer to Rasch estimations was determined. These 3 judges' appraisals were considered to establish the performance levels.
    Revista digital Matemática, Educación e Internet (http://www.tecdigital. itcr.ac.cr/revistamatematica/. 04/2013;
  • [Show abstract] [Hide abstract]
    ABSTRACT: In this article, we study the psychometric properties of a short scale (RL-14) created in order to ascertain the information / training received by employees and their psychophysiological disorders. The participants of this study were 1.489 employees in various sectors including education, construction, transport, health, industry, trade and hospitality. The results of an exploratory and confirmatory factor analysis show a structure consisting of two factors: "(F1) Information and Training" and "(F2) Psychophysiological Pathologies". In addition, the two factors obtained have adequate reliability and present evidence of the validity of the two factors if some sociodemographic variables, such as fatigue, burnout and job strain are taken as benchmarks. In short, this scale may be suitable for the proper identification of workplace hazards. Future research could use the RL-14 as a tool for screening in combination with other instruments.
    Anales de Psicología 01/2013; 29(1):217-224. · 0.55 Impact Factor

Full-text (2 Sources)

View
45 Downloads
Available from
May 31, 2014