ohe09nz81r的部落格

標題:

Statistics - t Test for a Correlation Coeffiicient

發問:

呢個t test 有無話n一定要小於或等於30先可以用呢~~? 如果個n=50 我可以點做(r=0.1029, alpha5%?要用咩formula? 更新: t 0.025,48=2.0106 更新 2: t 0.025,48=2.0106 how come? how to get it? from tables? my tables have no 48(df) ..any formulas for this?

• 橋口隆志的漫畫
• 邊到有土耳其雪糕食-
• 問1題數學題~

• 橋口隆志的漫畫
• 邊到有土耳其雪糕食-
• 問1題數學題~

 

此文章來自奇摩知識+如有不便請留言告知

最佳解答:

H0: p=0 H1 p≠0 critical region: tα/2,n-2 testing statistics: tcal= r *sqrt(n-2) / sqrt(1-r^2) if tcal >tα/2,n-2 ==> reject H0 .the sample correlation coffection can be used as an estimator of the population correlation coefficient. In this example, tcal= 0.1029 * sqrt(50-2)/sqart(1-0.1029^2) =0.7167 t 0.025,48=2.0106 Don't reject H0 You can also use F-test critical region : f1,n-2 testing statistics: Fcal=r^2(n-2)/(1-r^2) <======is t^2 2007-05-15 14:21:58 補充： (1) proportion:(In t-test table )dF: 40 48 60 2.021 x 2so (x-2.021)/(48-40)=(2-2.021)/(60-40) == x=2.0126(2) In excel type "=TINV(0.05,48)" 2007-05-15 14:23:04 補充： (2) In excel type "=TINV(0.05,48)"=2.01063358 <=prob=0.05 是因為excel 用2-tails 的,但table是1-tail(因為進位問題所以有些不同)

其他解答: