Comparing age groups

[seanjfox]

I have a set of customer data with an actual date of birth and from an external data set I have appended a modelled age banding/group. I have converted the actual dates of birth into the same range bandings/groups as those present in the modelled age bandings/groups. I would like to understand how accurate the modelled age banding is versus the actual age banding.

The age bandings are as follows:
18-25
26-35
36-45
46-55
56-65
66+

I have attached a sample of dummy data for reference

 

[Archidamus]

What are you asking? Are you conducting a linear regression and wondering if it makes sense to group the ages into these bands?

Well, I would do both. And based off the results determine what kind of model would work best.