It is a good idea to pool the 1970 form 1 metro sample and 1970 form 2 metro sample together

USA
1

Hi there,

I am trying to calculate house prices for all metropolitan areas. Therefore I would like to have as many observations as possible for each metropolitan areas in each year. Do you think it is plausible to pool the 1970 form 1 metro sample and 1970 form 2 metro sample together and use the weight (hhwt) to calculate the mean(median) house price for each metropolitan areas?

Thanks a lot.

All my best,

Vera

2

It is reasonable to combine the two 1970 metro samples in order to create a larger data set. This is possible because these samples are, for all practical purposes, mutually exclusive. See page 8 of the 1970 codebook for a discussion of this fact. Note, however, that the household weight (HHWT) is designed so that each of these samples, individually, represents the entire population. So, you will have to adjust the sampling weight in some way. A reasonable way to do this is to simply divide the weight variable by the number of samples you are pooling together. So, in your case you should generate a new weight variable that is equal to HHWT divided by 2.

1 Like
3

I saw some strange-looking numbers from this pooled sample, after laboring a bit I grew this hunch that I would need to exactly what you said. Just wanted to drop a reply thanking you for confirming this solution and saving me a headache!

1 Like