The command of cross-tabulation by using normalized weights?

[Lei Chai]

Hello everyone,
Does anyone know the command of cross-tabulation by using normalized weights?

Thanks in advance.

Lei

[Rich Goldstein]

not sure what you mean by "normalized" weights; note that "tabulate" allows 3 kinds of weights: fweights, aweights and iweights; are any of these the same as what you want? (if you're not sure what these are, type

help weight

to read up on these

[Lei Chai]

Hi Rich,

Thanks for your comment. Yes, I am going to use pweights. My original commands are:
svy [pw=WGHT_PER]
svy: tab maritalstatus age, col

But I was suggested to use normalized weights as well. I just do not know how to add normalized weights to my current commands?

Lei

[Steve Samuels]

There's no benefit to normalizing--it won't change the results of your svy: tab command, for example. And there are reasons not to.

The definition of a sampling, or probability, weight, is: \(w_i = \frac{1}{f_i}\), where \(f_i\) is the probability that sample member \(i\) was selected.

"Normalized" weights are probability weights that are scaled (divided by a constant) to sum to some constant C. I've seen C = 1 and C = sample size \(n\). This last was popular before the days of survey software. As I've said, this makes no difference to most analyses.

However there are good reasons not to normalize weights. If you do normalize:

1. You lose the possibility of estimating totals, including population counts.

2. If you use a finite population correction (fpc option to svyset), then estimates of the design effect DEFF be incorrect.

3. You lose interpretability. A subject's weight has the interpretation as the number of people in the population "represented" by the subject. This is easy to understand and it is often important to look at.

Note that the "probability" weights supplied with many data sets are not sampling weights, but sampling weights adjusted for non-response and post-stratified so that estimated sample totals match population totals known from other sources (like a Census). The reasons against normalizing apply to these weights as well.

Before posting again, please read FAQ 12 and learn how to put code, data listings, and results between CODE delimiters.