Compare two datasets and whether they agree

Question

I have two datasets and they both have the same set of independent variables:

9 of them are on scale from 0 till 100
3 of them are categorical(1 with two types categories, 1 with
three types of categories, 1 with    six types of categories)

The meaning of the dependent variables are the same but their scales are different.

How happy are you about your marriage? These scores are on a scale
from 0 till 5: Let's call this group A
Are happy with your marriage? These scores are as yes or no: Let's
call this group B

I want to check whether Group B and Group A perceived the Happiness differently by checking whether the combination (with interaction and without) of our independent variables effected the happiness   differently from each-other.

How can I test this?

Update:
Applying the assumption: a scale transformation: 1 till 3 = bad and 4 and 5 = good is possible for me.I want to merge the two groups together.

Dale C · Answer

You're probably going to need make assumptions on the dependents in A in order to map them onto the same scale as B (a very naive and simple example would be to convert all values above 3 from A to 1 and below 3 from A to 0, the assumption being these values correspond to the same in B etc.)

Then you're just comparing proportions of similarly scaled categorical variables, which can easily be done with a $chi^2$ test.

Kamil Kaczmarek · Answer

I think, that you cannot. Comparing data requires stating objective, numerical differences, which you don't have. You would have to assume, that answers 'yes' or 'no' corresponds to some numbers, ranges of numbers or some combination of numbers. But that would be hard simplification only based on your subjective assumption. People perception of these two scales may be so different, that ascribing any numbers would be impossible.

Answered by Kamil Kaczmarek on November 9, 2021

Compare two datasets and whether they agree

2 Answers

Add your own answers!

Ask a Question