If my weights shrink the number of observations from 160 to 26, can I still report row percentages?


Please can someone help me?

I am trying to put together an estimate of the number of Sikhs who were insulted on grounds of their religion. I am using the Understanding Society dataset (a uk household panel).

I've created a table with religion and a binary variable for having been insulted or not across the side. The weights are shrinking the number of observations from 160 to 26. It it still safe to report the row percentages in this case?

Any help would be much appreciated. Thanks,

Richard Norrie