Finding correlation between size, weight and location in a network graph

I have an undirected graph of students, their friendship ratings, and amount of effort paid in class in the following format:

Person 1(Student ID)  Person 2(Student ID)  Strength    Class_Room  Effort_Paid
A                     B                     5           112         1
G                     A                     2           112         4
K                     E                     3           113         2


I was told that certain classes have better atmosphere between students which results in more effort paid in class, and I am trying to mathematically verify that claim. Just to be clear, my question is - are there any relationship between network size, friendship strength (edge weight), effort paid based on classroom breakdown?

So far I've tried using the average strength and average number of connections for each classroom but found no significant correlation with effort paid. Those are pretty basic items to calculate and correlate, I suspect a more sophisticated model would help me unlock some insights. I've been reading into network diffusion and centrality measure but I don't think either of them apply in this case. Do you have any idea which model could potentially benefit me?