2013-07-06 11:40:02解讀統計與研究譯者

ICC在SPSS的用法

使用SPSS計算ICC時 會詢問研究者使用哪一種ANOVA模式 接下來要選擇consistency或absolute agreement 不同的選擇很可能產生不同的結果而且差異不小

有三種ANOVA模式:

(1)One-way random:只考慮對象效果

(2)Two-way random:同時考慮評分者與對象效果

(3)Two-way mixed:與(2)差在評分者效果限於眼下研究所選之評分者

那麼到底要選擇consistency或absolute agreement?

(1)Consistency: Rater variability沒有放進分母

(2)Absolute agreement: Rater variability有放進分母

ICC理論公式如[1]: source: http://www.john-uebersax.com/stat/icc.htm

                                     s 2(b)    
ICC    =   ------------   [1] 

  

s 2(b) + s 2 (w)

 

Note. (w) within subjects, (b) between subjects.

 

再看一下分類: source: https://en.wikipedia.org/wiki/Intraclass_correlation

Shrout and Fleiss six modelsDescription in SPSS
ICC(1,1)One-way random single measures
ICC(1,n)One-way random average measures
ICC(2,1)Two-way random single measures (Consistency/Absolute agreement)
ICC(2,n)Two-way random average measures (Consistency/Absolute agreement)
ICC(3,1)Two-way mixed single measures (Consistency/Absolute agreement)
ICC(3,n)Two-way mixed average measures (Consistency/Absolute agreement)
單單 2013-07-26 17:46:34

非常有用的整理! 解決得很久的疑慮! 感謝分享

版主回應
把愛傳下去! 2013-07-27 12:40:50