I have recently posted a question but I might have made it too complicated to gain any response. I try to simplify my question and hopefully I can get some insight from the experts here.

I have interviewed 3 members (father, mother and child) in a family. Each of them rate the level of permissiveness of mother in a 5-item questionnaire in 5-point likert scale. I would like to see the inter-rater reliability among these 3 members.

I read about intraclass correlations (ICC) but it seems that it is not directly related to my case, because I only have 1 subject in my study. I am considering if I can treat the 5 items in the questionnaire as 5 subjects and conduct the test. I make this adjustment as I do not know any alternative statistical way and the five items belong to the same construct.

In this case, my data base looks like this:
...........father.....mother....child
item 1.....1............2...........2
item 2.....2............2...........1
item 3.....5............3...........4
item 4.....2............1...........1
item 5.....2............2...........2

I don't know if I can analyse my data this way or if it makes any sense statistically. I hope you can share your thoughts or tell me any alternative ways to achieve this task. (I use SPSS)

Millions of thanks!!