Normalizing Interconnected Data: A-A, A-B, B-B

[Spiderman]

This may be a simple problem, but I wanted to run it by some other people before using my solution.

I have two distinct sets of items A and B which may or may not be connected to one another. I want to know whether or not the interconnections between them are significiantly different, i.e are the number of connections between A-A, A-B, and B-B different - are A's connected more to A's, for example. However, there are many more B's than A's. Normally I would just divide the value by the number of items, but how do I do this with interconnected items? Is the normalized value of the number of A-A connections =

number of connections/(A*A)

And similarly for A-B: number of connections/(A*B)

I can't determine if this is right or not.

 

[Rasine]

you may want to do a chi squared goodness of fit test. seeing if a-a matches the data for a-b..ect

 

[tacman]

It's always a good idea to seek feedback and input from others before implementing a solution, so kudos to you for doing so! In terms of normalizing interconnected data, your approach seems to be on the right track. However, I would suggest using a different formula for calculating the normalized value.

Instead of dividing by the product of the number of A's and B's, I would recommend dividing by the total number of possible connections. In this case, the total number of possible connections would be (A + B)^2, since you are looking at all possible pairs of A's and B's, including A-A, A-B, B-A, and B-B.

So the normalized value for A-A connections would be: number of A-A connections / (A+B)^2

And the normalized value for A-B connections would be: number of A-B connections / (A+B)^2

This formula takes into account the fact that there are more B's than A's, and ensures that the normalized values are comparable across different sets of interconnected data.

I hope this helps and good luck with your data analysis!