Quantitative Research in Finance: An Analysis of Statistics Techniques Used by Articles Published in Qualified Scientific Journals from 2007 to 2009

Abstract

This study aimed to examine whether there are differences, in terms of sophistication and accuracy, in statistical techniques used by finance articles published in journals classified by CAPES, from 2007 to 2009. The sample consisted of 100 articles from 40 journals, national and international, inserted between strata A1 and B5. We used descriptive statistics in order to characterize the sample, and the chi-square test of independence, by means of which was examined the degree of independence between variables. The results suggest that the higher the classification level of the stratum, the greater the sophistication of statistical techniques employed in the articles. However, regarding the accuracy in applying the techniques, there were no significant differences between one stratum and another. Most of the articles analyzed (57%) not validated or partially validated the assumptions required for their use. Thus, it was realized that there was not proper concern with this aspect, even in well qualified strata. We emphasize that the results are restricted to the sample used. We recommend that future studies use large samples, including other means of publication, like events, both in finance as in other areas of knowledge.