The analysis of 250,000 p-values published in 8 major psychology journals from 1985 to 2013 revealed half of them were inconsistent with their test statistic that may affect test conclusions and that inconsistencies were more frequent for significant p-values.

See the source article by following the link below:
The analysis of 250,000 p-values published in 8 major psychology journals from 1985 to 2013 revealed half of them were inconsistent with their test statistic that may affect test conclusions and that inconsistencies were more frequent for significant p-values.