Lastly, the Monte Carlo methods showed that the Mann‐
Whitney U test can give wrongfully significant results, that
is to say the erroneous acceptance of the alternative
hypothesis (Robert & Casella, 2004). This type of results is at
risk to be obtained whenever one’s samples are drawn from
two populations with a same average but with different
variances. In this type of situations, it is largely more reliable
to use the t‐test which gives a possibility for the samples to
come from distributions with different variances. The alpha
(α) error or of type I is to reject H0 whereas this one is true.
This error is thus amplified when Mann‐Whitney U is
applied in a situation of heteroscedasticity or distinct
variances. In addition, some solutions exist to this major
problem (see Kasuya, 2001).