It has been over 10 years that I implemented the RCI fucntion into scan.
I reread the original paper of Hageman and came to the conclusion that I would not recommend to use the present implementation of the Hageman index in scan.
Hageman estimates the reliability of the differenc scores and therefore needs a correlation of pre and post scores of a lot of cases. Which you usually don't have in single-case designs. I simply ignored this aspect in scan which was not a good solution.
The original Jacobson or the improvement by Christensen had single-case designs in mind and are more appropriate (and straight forward).
I will implemenet a reworked version of the rci function in one of the next scan updates.
Cheers,
Jürgen
It has been over 10 years that I implemented the RCI fucntion into scan. I reread the original paper of Hageman and came to the conclusion that I would not recommend to use the present implementation of the Hageman index in scan. Hageman estimates the reliability of the differenc scores and therefore needs a correlation of pre and post scores of a lot of cases. Which you usually don't have in single-case designs. I simply ignored this aspect in scan which was not a good solution. The original Jacobson or the improvement by Christensen had single-case designs in mind and are more appropriate (and straight forward). I will implemenet a reworked version of the rci function in one of the next scan updates. Cheers, Jürgen
Originally posted by @jazznbass in https://github.com/jazznbass/scan/discussions/140#discussioncomment-9210375