Hello,

I have question about data preparation.
I have group comparison data and also correlation data, how can I include correlation data for analysis. should I put t-value as + for positive correlation and put - for negative correlation?

Thank you.

Sang Soo

Dear Sang Soo,

You can convert correlation coefficients into t-values, i.e.,

t = r * sqrt( (n-2) / (1-r^2) )

Hope this helps,

Joaquim

Hello, Joaquim
for the parametric analysis, usually we have already t-value for the table. I wonder how should I differenciate  t-value for negative or positive correlation... 
Do you have any suggestion?
 
Thank you for your help.
Sang Soo. 

Dear Sang,

t-values for positive correlations should be >0, whereas t-values for negative correlations should be <0.

Hope this helps,

Joaquim

Thanks, Dr. Radua!

I have a question regarding to this. Then how should I set n in the sdm_table?

Just put n (one column) as it's one-sample t-test?

Best,

Chao-Gan

------
Dear Sang Soo,

You can convert correlation coefficients into t-values, i.e.,
t = r * sqrt( (n-2) / (1-r^2) )
Hope this helps,
Joaquim

In addition, how should I dear with F values?

Dear Chao-Gan, this seems reasonable.

Dear Chao-Gan, I think that a simple solution might be to use t = sqrt(F)

> Thanks, Joaquim!