Errores de interpretación de los métodos estadísticos: importancia y recomendaciones

Psicothema, ISSN 0214-9915, Vol. 18, Nº. 4, 2006, pags. 848-856 01/2006;
Source: OAI

ABSTRACT Trabajos empíricos previos han identificado las opiniones de los investigadores respecto de las pruebas de significación y de otros recursos estadísticos, algunas de estas opiniones resultan totalmente inaceptables. En esta investigación comprobamos el grado en que estos errores aparecen en una muestra española de profesores e investigadores de universidad mediante la aplicación de un cuestionario. Los datos obtenidos son importantes para: a) la prevención de interpretaciones inadecuadas; b) la corrección de usos incorrectos; c) el análisis de alternativas posibles; y d) proponer cambios editoriales en los criterios de publicación.

  • Source
    [Show abstract] [Hide abstract]
    ABSTRACT: The aim of this study is to compare partially disaggregated structural models, using item parcels as indicators of latent variables, and path analysis models, assuming totally aggregated items make up measurements without errors of the constructs. Data come from two empirical studies: one focused on education and another one focused on health, in such a way we can prove the constructs evaluated do not interfere with methodological results showed. The results in both studies are consistent. On one hand, the use of items parcels as indicators of the constructs rather than path models increase the value of the parameters in absolute value, and residual variance is reduced. This is reflected in an increase of the explained variance by the models. Parcel construction method does not seem to be a relevant factor. On the other hand, goodness-of-fit indexes of path models show a better fit. Summarizing, latent variable models would be better fitted to reality, improving the percentage of explained variance, but losing fit due to the higher number of parameters to be estimated.
    International Journal of Psychology 01/2009; 2(2):91-110. · 0.40 Impact Factor
  • [Show abstract] [Hide abstract]
    ABSTRACT: The generation of scientific knowledge in Psychology has made significant headway over the last decades, as the number of articles published in high impact journals has risen substantially. Breakthroughs in our understanding of the phenomena under study demand a better theoretical elaboration of work hypotheses, efficient application of research designs, and special rigour concerning the use of statistical methodology. Anyway, a rise in productivity does not always mean the achievement of high scientific standards. On the whole, statistical use may entail a source of negative effects on the quality of research, both due to (1) the degree of difficulty inherent to some methods to be understood and applied and (2) the commission of a series of errors and mainly the omission of key information needed to assess the adequacy of the analyses carried out. Despite the existence of noteworthy studies in the literature aimed at criticising these misuses (published specifically as improvement guides), the occurrence of statistical malpractice has to be overcome. Given the growing complexity of theories put forward in Psychology in general and in Clinical and Health Psychology in particular, the likelihood of these errors has increased. Therefore, the primary aim of this work is to provide a set of key statistical recommendations for authors to apply appropriate standards of methodological rigour, and for reviewers to be firm when it comes to demanding a series of sine qua non conditions for the publication of papers.
    Clínica y Salud. 03/2013; 24(1):47-54.
  • [Show abstract] [Hide abstract]
    ABSTRACT: A programme of educational innovation was implemented to improve the attitude toward research methods for psychology students. Students used real data to test their own hypotheses using different statistical methods. Specifically, they implemented two key methodological principles in research methods: sample size and ordinal claims (Frick, 1996). Students had to work in teams, using Internet data bases for developing the theoretical background of the study, the definition of the objectives, the definition of variables, data analysis, and discussion of results. We assessed the attitudes towards research methods before and after the study. Results showed that: 1) the students improved their skills in research methods, and 2) the attitudes towards methodology were more positive.
    Escritos de Psicología (Internet). 12/2010; 3(4):1-10.

Full-text (2 Sources)

Available from
May 28, 2014