Intraclass Correlation Coefficient (ICC)

[seol]

Dear Marju

Thanks for your suggestions again. Although I really would like to install MFRM model into jamovi environment, I can not find useful R package until now. If you find useful R package for MFRM model, please let me know.

Best Regards
Seol

[lauramar]

Hi Seol,

I work in the field of language assessment and have very limited knowledge in R. I found one tutorial on rpubs (https://rpubs.com/isbell_daniel/735520) where MFRM was tried out with R. Is this useful?

Best

Marju

[lauramar]

Hi Seol
I have a few questions about the MFRM module to make sure I'm using it correctly.
First, scores: I read that in case of TAM (I understand this is the package the MFRM module is based on), the lowest score on the scale should be 0. So, if in practice the lowest score I can give for a performance is 1, in the dataset I should replace it with 0 etc for the other scores. Can you confirm whether this is correct or it doesn't matter whether the lowest score is 0 or 1?
Second, the task facet: my understanding is that the most difficult task should have the highest logit value. In the dataset that I used, the tasks seem to be ordered the other way around, the one for which raters can give the highest score seems to be the easiest task. I will include the dataset just in case you can find time to look at it. The scores start from 1 and there are 7 tasks. Can you help me understand how to interpret the results for the tasks?
Many thanks
Marju

[seol]

Hi
Thanks for using MFRM model in snowIRT module.
I'm not helping with statistical interpretation, but here's the MFRM model reference I used to develop the module. I hope you find this helpful.

https://bookdown.org/chua/new_rasch_dem ... model.html

Best Regards
Seol