The sample size is to small to be statistically relevant.
Not true, any sample size over 32 is relevant.