My organization has a policy that any survey response with a sample size LT 6 is not reported. This is applied to the answer, not the whole question, e.g. in the "state" example below we would suppress the response "Texas" and in the "gender" example, we would suppress "male."
My question is: is this the right way to be doing this or should we be applying sample size rules to the total respondents of a question (e.g. the 23 "state" variable and 12 in the "gender" variable). To suppress just the category doesn't make a lot of sense to me. E.g., with gender, you can suppress the "male" response -- but the percent female is based on the percent male so it doesn't make a lot of sense.
Sorry if this is a dumb question but appreciate any help out there!
Variable "state" with unweighted frequencies California: 9 Texas: 4 Other: 10 Total: 23
Variable "gender" with unweighted frequencies Male: 2 Female: 10 Total: 12