我的论文主题是焦虑的干预研究，之前进行了干预之前和之后焦虑得分的对比，使用了重复测量的方差分析，也做了性别前后测焦虑得分的重复测量的方差分析。审稿人意见如下。
Since anxiety scores are unlikely to achieve ‘ratio’ level of measurement, using non-parametric methods [non-parametric parallel of independent samples ‘t’ test is Mann-Whitney ‘U’ test] should have been (expected to be) done.
Comparing post test scores by ANCOVA where pre test scores taken as covariate is straight forward compare to mixed factorial ANOVA. For ‘Gender’ interaction one can use ‘interaction test’ then. What is done is not incorrect. It is correct but the ‘round-about’ way.

请问审稿人的意见到底是什么意思？ 是不是必须使用非参数检验？为什么重复测量的方差分析不合适？而我之前的分析不能说错，是否意味着可以保留，然后加上非参检验？ 返回小木虫查看更多