Covariance of two variables, multiple individuals

[itendswithe]

I have a data set with two variables (A and B), but variable B was measured for multiple individuals (B1, B2, B3, B4). Is there a way to calculate a single covariance (and, at that, sum of cross products) for my data? Or, does the nature of the data dictate a separate covariance calculation for each individual?


Thanks!

 

[hlsmith]

Please describe these data in more detail. Currently your description leads me to think variable A consists of one person (no idea how many observations) and variable B with multiple people (no idea how many people) ? And is the one person in variable A in variable B?

See the vagueness!

 

[itendswithe]

I totally understand the confusion! Sorry for not being clearer. I think an example might be easier. Say I have 4 individuals assess their anxiety levels (variable B) daily for 10 days, and I also record the rainfall for these 10 days (variable A), and I want to examine the relationship between rainfall and anxiety, so my data looks a bit like the attached file (which I just made up numbers for).

Does that make it clearer?
Thanks!

 

[hlsmith]

It may be more complex than you want but this is perfect for a multilevel model (hierarchical regression).