On 2016 Dec 13, UFRJ Neurobiology and Reproducibility Journal Club commented:

There are also issues in the statistical analysis, as a large number of behavioral measures are obtained, with relatively few significant results, which due to the amount of statistical comparisons (more than 20 in total) may well be just spurious effects. However, the authors never seem to discuss this possibility. Moreover, in the passive avoidance task the results mention Z scores and chi-square results, while the figure and methods mention that the analysis was performed with Mann-Whitney and Kruskal-Wallis tests, which should not yield either chi-square or Z scores. Finally, the conclusion of the authors that testosterone and DHT have different effects is not warranted, as even though each drug shows significant differences against placebo in different tests, in none of them a difference between both treatments was found. Moreover, the improvement in the water maze retention test in the testosterone group seems to have been inferred from intra-group comparisons between quadrants in these groups, but no comparison between groups was performed. The assumption of a difference between treatments is thus erroneous, as the statistical analysis does not compare the groups directly (for more information on this common statistical error, see Nieuwenhuis S, 2011).

On 2016 Dec 13, UFRJ Neurobiology and Reproducibility Journal Club commented:

There are also issues in the statistical analysis, as a large number of behavioral measures are obtained, with relatively few significant results, which due to the amount of statistical comparisons (more than 20 in total) may well be just spurious effects. However, the authors never seem to discuss this possibility. Moreover, in the passive avoidance task the results mention Z scores and chi-square results, while the figure and methods mention that the analysis was performed with Mann-Whitney and Kruskal-Wallis tests, which should not yield either chi-square or Z scores. Finally, the conclusion of the authors that testosterone and DHT have different effects is not warranted, as even though each drug shows significant differences against placebo in different tests, in none of them a difference between both treatments was found. Moreover, the improvement in the water maze retention test in the testosterone group seems to have been inferred from intra-group comparisons between quadrants in these groups, but no comparison between groups was performed. The assumption of a difference between treatments is thus erroneous, as the statistical analysis does not compare the groups directly (for more information on this common statistical error, see Nieuwenhuis S, 2011).