Resumo

PONCE RENOVA, Héctor Francisco; CERVANTES ARREOLA, Diana Irasema  e  ANGUIANO ESCOBAR, Beatriz. Quality Analysis for Educational Articles with Experimental Designs. RIDE. Rev. Iberoam. Investig. Desarro. Educ [online]. 2021, vol.12, n.23, e03.  Epub 18-Out-2021. ISSN 2007-7467.  https://doi.org/10.23913/ride.v12i23.981.

The objective of this work was to evaluate the statistical quality of refereed articles (n = 34) with experimental designs and to list a series of recommendations to increase their quality. The articles were obtained from the Red Iberoamericana de Innovación y Conocimiento Científico (Redib) with the selection criteria: a) having carried out an experimental design in educational research and b) having used a parametric analysis. Under a quantitative, exploratory, descriptive, and inferential statistics methodology, the quality of the articles was evaluated under four criteria: statistical power, susceptibility to replication, access to databases to verify the results and statistics suggested by the American Psychological Association (APA). In the results, it was found that most of the articles did not give enough information to support their conclusions and the reporting of the statistics had statistically significant differences and showed a large effect size. Therefore, they were evaluated with poor quality. In conclusion, and to improve quality, it is recommended to use guides for statistical analysis, replication of studies and to give access to databases so that other authors can observe what was done.

Palavras-chave : quality; experimental design; inference; power; replication.