Re: error message after weighting : At least one case in the data file had a weight value that was user-missing, system-missing, zero, or negative.

[Frans Houweling]

The freq table is weighted. Use SHOW WEIGHT to see the weighing variable, and WEIGHT OFF to reset.

On 8/24/24 16:48, ft gmail wrote:

Hi,

I try to understand what the error message printed below means, and what to do.

In fact, I analyse a data set from the European Social Survey which is provided with a weight variable anweight which I used.

I got the error message:

warning: FREQUENCIES: At least one case in the data file had a weight value that was user-missing, system-missing, zero, or negative.  These case(s) were ignored.
    1 | freq anweight.
      |              ^

A frequency output of anweight shows no user- or system-missing, negative or zero value. For a multi-million representative and publicly downloadable bi-annual population study this would be surprising.

             Statistics
╭─────────┬────────────────────────╮
│         │anweight Analysis weight│
├─────────┼────────────────────────┤
│N Valid  │                42359,68│
│  Missing│                     ,00│
├─────────┼────────────────────────┤
│Mean     │                    3,85│
├─────────┼────────────────────────┤
│Std Dev  │                    2,41│
├─────────┼────────────────────────┤
│Minimum  │                     ,12│
├─────────┼────────────────────────┤
│Maximum  │                    22,0│
╰─────────┴────────────────────────╯

Any ideas ?

Thanks in advance

- ftr