Bootstrapping for weighted proportions (non-integers)

Moderators: statman, Analyst Techy, andris, Fierce, GerineL, Smash

Posts: 1
Joined: Tue May 03, 2016 3:36 am

Bootstrapping for weighted proportions (non-integers)

Postby pungentprincess » Tue May 03, 2016 3:45 am

Hello all!

I'm trying to figure out how to perform a bootstrap analysis on weighted proportions. Have a binary outcome (Y/N) for different groups across different countries.

SPSS tells me that in order for me to perform bootstrap analysis to get confidence intervals, I need my weights to be integers. I tried my weights by a factor of 10,000, so that I would get rid of the decimals and convert the weights into whole numbers.

When I re-ran the bootstrapping analysis, weighted by my new weight variable, (done through the crosstabs option), I get the exact same result - SPSS tells me that non-integer weights can't be used in bootstrapping analysis.

I tried a few things to try to make it work. I adjusted the newly converted weight variable to make sure there were no decimal places (in the decimal tab under variables). I also tried changing my variable from a scale variable to a nominal variable (also under the variables tab).

How could I get this to work? Or perhaps there is another option to get confidence intervals for weighted proportions? Thank you in advance!

Who is online

Users browsing this forum: No registered users and 2 guests