Результаты (
английский) 3:
[копия]Скопировано!
a distribution free procedure to perform such a test is the bootstrap procedure.
i, in this procedure you draw many random samples out of
the real sample with replacement. it is important that these random samples. all have the same sample size as a real sample. for every random
sample compute the corresponding statistic (here the mean) and store it.to test your hypotheses compute the percentile of the hypothesized value
(10, 100) in the resulting vector. this percentile corresponds statistics to
a p value in an usual statistical test.
переводится, пожалуйста, подождите..
