Caperos and Pardo reviewed the statistical results reported in 186 articles published in four Spanish psychology journals with regard to consistency errors, i.e. p-values that do not correspond to the reported value of a tested statistic and its degrees of freedom.
Of the reviewed articles, 102 contained at least one of the statistics selected for the study: Fisher- F , Student- t and Pearson-c2. Of the 1212 statistics reviewed, 12.2% represented a consistency error. In 2.3% of the cases, the correct p-value would have led to a different conclusion. Of 102 articles, 48% had at least one consistency error, and in 17.6% of the articles the correct p-value would have changed at least one conclusion.
The authors conclude that consistency errors are substantial and that Spanish psychology journals need to improve the quality and precision of the statistical results.
Caperos JM, Pardo A. Consistency errors in p-values reported in Spanish psychology journals. Psicothema 2013;25:408-414.