I have a dataset where each observation is a residential unit. The units are observed on two characteristics

$$\mathbf x_i =(x_{1i},x_{2i}),$$

where the first is the size of the residential unit measured in square meters and the second is the number of rooms in the residential unit.

Some observations have a missing value for the size and some for the number of rooms. I would like to impute the missing values, so my question is:

What method can I use to impute the missing values?

I do not have much experience in data imputation, but I was thinking of simply using the observations where neither variable is missing to construct

(1) A count model for the number of rooms as a function of the size and

(2) A regression,median regression or perhaps non-linear regression model for the size as a function the number of rooms

and then simply use the prediction from these models to impute the missing values. So my follow up question is

Would that be a way to do it, or would I be making some kind of rookie mistake according to standard imputation methods?

Any guidance is appreciated.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.