楼主: chyshl
9067 9

[一般统计问题] Stata的anova命令没有多重比较的选项吗? [推广有奖]

  • 0关注
  • 46粉丝

院士

7%

还不是VIP/贵宾

-

威望
0
论坛币
1766 个
通用积分
149.2889
学术水平
130 点
热心指数
153 点
信用等级
102 点
经验
332097 点
帖子
3179
精华
0
在线时间
2894 小时
注册时间
2005-6-5
最后登录
2022-10-15

楼主
chyshl 发表于 2009-7-28 23:57:30 |AI写论文

+2 论坛币
k人 参与回答

经管之家送您一份

应届毕业生专属福利!

求职就业群
赵安豆老师微信:zhaoandou666

经管之家联合CDA

送您一个全额奖学金名额~ !

感谢您参与论坛问题回答

经管之家送您两个论坛币!

+2 论坛币
听说Stata不如SPSS的地方之一就是它的方差分析功能较弱,比如,SPSS的方差分析功能(不包括oneway)后有一系列多重比较选项,当然最常用的是S-N-K啦,而Stata就只能给出一个总体有无差异的结论,俺找了半天,貌似Stata的anona命令下没有多重比较功能哦。
到底有没有?大牛们解释一下?如果需要自编命令,如何实现?
二维码

扫码加我 拉你入群

请注明:姓名-公司-职位

以便审核进群资格,未注明则拒绝

关键词:Stata ANOVA anov tata 多重比较 Stata 命令 选项 ANOVA

沙发
chyshl 发表于 2009-7-29 00:01:36
不好意思,正文中打错了一个字母:anona应该为anova.

藤椅
biostat 发表于 2009-7-29 10:45:53
help oneway                                                     dialog:  oneway
-------------------------------------------------------------------------------

Title

    [R] oneway -- One-way analysis of variance


Syntax

        oneway response_var factor_var [if] [in] [weight] [, options]

    options           description
    -------------------------------------------------------------------------
    Main
      bonferroni      Bonferroni multiple-comparison test
      scheffe         Scheffe multiple-comparison test
      sidak           Sidak multiple-comparison test
      tabulate        produce summary table
      [no]means       include or suppress means; default is means
      [no]standard    include or suppress standard deviations; default is
                        standard
      [no]freq        include or suppress frequencies; default is freq
      [no]obs         include or suppress number of obs; default is obs if
                        data are weighted
      noanova         suppress the ANOVA table
      nolabel         show numeric codes, not labels
      wrap            do not break wide tables
      missing         treat missing values as categories
    -------------------------------------------------------------------------
    by is allowed; see [D] by.
    aweights and fweights are allowed; see weight.


Description

    The oneway command reports one-way analysis-of-variance (ANOVA) models
    and performs multiple-comparison tests.

    If you wish to fit more complicated ANOVA layouts or wish to fit
    analysis-of-covariance (ANOCOVA), see [R] anova.

    See [D] encode for examples of fitting ANOVA models on string variables.

    See [R] loneway for an alternative oneway with slightly different
    features.


Options

        +------+
    ----+ Main +-------------------------------------------------------------

    bonferroni reports the results of a Bonferroni multiple-comparison test.

    scheffe reports the results of a Scheffe multiple-comparison test.

    sidak reports the results of a Sidak multiple-comparison test.

    tabulate produces a table of summary statistics of the response_var by
        levels of the factor_var.  The table includes the mean, standard
        deviation, frequency, and, if the data are weighted, the number of
        observations.  Individual elements of the table may be included or
        suppressed by the [no]means, [no]standard, [no]freq, and [no]obs
        options.  For example, typing

            . oneway response factor, tabulate means standard

        produces a summary table that contains only the means and standard
        deviations.  You could achieve the same result by typing

            . oneway response factor, tabulate nofreq

    [no]means includes or suppresses only the means from the table produced
        by the tabulate option.  See tabulate above.

    [no]standard includes or suppresses only the standard deviation from the
        table produced by the tabulate option.  See tabulate above.

    [no]freq includes or suppresses only the frequencies from the table
        produced by the tabulate option.  See tabulate above.

    [no]obs includes or suppresses only the reported number of observations
        from the table produced by the tabulate option.  If the data are not
        weighted, only the frequency is reported.  If the data are weighted,
        the frequency refers to the sum of the weights.  See tabulate above.

    noanova suppresses the display of the ANOVA table.

    nolabel causes the numeric codes to be displayed rather than the value
        labels in the ANOVA and multiple-comparison test tables.

    wrap requests that Stata not break up wide tables to make them readable.

    missing requests that missing values of factor_var be treated as a
        category rather than as observations to be omitted from the analysis.


Examples

    ---------------------------------------------------------------------------
    Setup
        . webuse apple
        . oneway weight treatment

    Obtaining observed means
        . oneway weight treatment, tabulate

    Bonferroni multiple-comparison test
        . oneway weight treatment, bonferroni

    Scheffe multiple-comparison test
        . oneway weight treatment, scheffe

    ---------------------------------------------------------------------------
    Setup
        . webuse census8

    With weighted data
        . oneway drate region [w=pop]
    ---------------------------------------------------------------------------


Saved results

    oneway saves the following in r():

    Scalars   
      r(N)           number of observations
      r(F)           F statistic
      r(df_r)        within-group degrees of freedom
      r(mss)         between-group sum of squares
      r(df_m)        between-group degrees of freedom
      r(rss)         within-group sum of squares
      r(chi2bart)    Bartlett's chi-squared
      r(df_bart)     Bartlett's degrees of freedom


Also see

    Manual:  [R] oneway

    Online:  [R] anova, [R] loneway

板凳
biostat 发表于 2009-7-29 10:46:43
可见提供了三种两两比较的方法:
  bonferroni      Bonferroni multiple-comparison test
      scheffe         Scheffe multiple-comparison test
      sidak           Sidak multiple-comparison test

报纸
kkwei 发表于 2009-7-29 19:12:27
比较而言,spss的方差分析能力更强大……

地板
eblog 发表于 2009-7-29 20:19:15
chyshl 发表于 2009-7-28 23:57
听说Stata不如SPSS的地方之一就是它的方差分析功能较弱,比如,SPSS的方差分析功能(不包括oneway)后有一系列多重比较选项,当然最常用的是S-N-K啦,而Stata就只能给出一个总体有无差异的结论,俺找了半天,貌似Stata的anona命令下没有多重比较功能哦。
到底有没有?大牛们解释一下?如果需要自编命令,如何实现?
你说的是多重比较检验吧?Stata只内置了三种:
      bonferroni      Bonferroni multiple-comparison test
      scheffe         Scheffe multiple-comparison test
      sidak           Sidak multiple-comparison test

如果想使用LSD、S-N-K其他检验方法可以考虑使用Spss软件

7
chyshl 发表于 2009-7-29 21:37:46
学习了,谢谢!
不过楼上弟兄说的都是oneway下的方法耶,两因素或多因素anova下还是没找到啊。

8
eblog 发表于 2009-7-29 22:45:27
chyshl 发表于 2009-7-29 21:37
学习了,谢谢!
不过楼上弟兄说的都是oneway下的方法耶,两因素或多因素anova下还是没找到啊。
help anova postestimation

9
chyshl 发表于 2009-8-1 00:02:09
哈!找到啦,感谢!

10
Jerel 发表于 2013-5-3 10:24:22
还是没看明白

您需要登录后才可以回帖 登录 | 我要注册

本版微信群
加好友,备注jltj
拉您入交流群
GMT+8, 2025-12-24 12:10