This post provides directions on calculating sample size.
One, from the IBM survey, start with Question 19 from the sample instead of Question 18 to calculate the population mean, and standard deviation. Be prepared to explain why you use Question 19 instead Question 18.
Two, for D, or level of precision, use +/- 5. That is, you want to the sample to be within 5 points of the population mean.
Three, plug-in your values for D (5), σ (calculated in step one), and z (1.96) into the formula n = (σ²*z²) / D² to calculate your sample size.
Four, if your sample size is greater than 10% of the population, then you need to correct with the formulae nc = (n*N) / (N+n+-1) where n equals the size of your needed sample and N equals the size of the population.
Remember, your population depends on your unit level of analysis.