ICC QUESTION - intraclass correlation coefficient of questionaire

Hi fellows,
I am trying to check the reliability of a translated questionnaire.
I made each participant to answer 12 questions * 2 languages.
I ran ICC tests for each question and got nice numbers.

My question is how to present it in an article. Now I have 12 different ICC scores. In similar articles it is always mentioned only one general ICC for the whole questionnaire. Should I just do a simple average of all the ICC values I've got?

Thanks!