a significance level of .25 may be more appropriate for deciding whether to do more research. Remember that the significance level chosen and the consequences of a Type I or a Type II error are determined by what the results will be used for.