Repeated measures ANOVA post hoc testing Bonferroni correction

Question

bsriv 2023-2-2

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1905396-repeated-measures-anova-post-hoc-testing-bonferroni-correction

回答： Koen van der Laan 2023-11-8

Hi, I am using a repeated measures ANOVA, and I would like to use the multcompare function to do Bonferroni correction.

When I use the default Tukey's HSD it works

multcompare(rm,'instruction')
ans =
  2×7 table
    instruction_1    instruction_2    Difference    StdErr      pValue      Lower      Upper  
    _____________    _____________    __________    _______    ________    _______    ________
          1                2            2.0082      0.85855    0.029293    0.22279      3.7937
          2                1           -2.0082      0.85855    0.029293    -3.7937    -0.22279

But when I try to use the bonferroni correction it doesn't

          multcompare(rm,'instruction',"CriticalValueType","bonferroni")
Error using internal.stats.parseArgs
Invalid parameter name: CriticalValueType.
Error in RepeatedMeasuresModel/multcompare (line 1132)
                internal.stats.parseArgs(okargs,defaults,varargin{:});

I'm not sure what the issue with my syntax is- please help!

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

Rik 2023-2-2

在 MATLAB Online 中打开

Well, your original code already results in an error for me. Without your initial code working, it is not possible to say why your second doesn't work.

load carsmall
[p,t,stats] = anova1(MPG,Origin);clc,close all
multcompare(stats,'instruction')
Error using internal.stats.parseArgs
Wrong number of arguments.

Error in multcompare (line 121)
        internal.stats.parseArgs(okargs,defaults,varargin{:});

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Koen van der Laan 2023-11-8

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/1905396-repeated-measures-anova-post-hoc-testing-bonferroni-correction#answer_1348742

Hello,

I recently experienced the same problem and found a answer. you need to use 'ComparisonType' to set the correction method when using a repeated measures model as input for 'multcompare'.

see this page:

https://nl.mathworks.com/help/stats/repeatedmeasuresmodel.multcompare.html#namevaluepairarguments