Select your confidence interval and call this "C." The confidence interval is the range within which the true proportion is expected to fall. For example, if you wanted the range to be within 3 percent above or below the percentage from your survey, you would use 0.03 for C.

Select your confidence level. This is the percentage of the time that the true proportion will lie within your confidence interval. The more important the study, the higher the confidence level. For example, a medical study may require a 99 percent confidence level, while a poll for a local election may only desire a 90 percent confidence level.

Convert your confidence level into a z-score, using the z-score chart, and call it "Z." For example, a 99 percent confidence interval would result in a z-score of 2.58.

Estimate the percentage of people who will select the majority option and call this "P." For example, if you expect 58 percent of the people to vote for the Democratic candidate, you would use 0.58 for P.

Plug your values for C, Z and P into the following equation to determine how large you need your sample size to be: (Z^2 * P * (1 - P))/C^2. For example, if you had a z-score of 2.58, a percentage of 0.58 and a confidence interval of 0.03, you would plug those numbers in to make your expression (2.58^2_0.58_(1-0.58))/0.03^2, which comes out to be 1801.67, meaning your sample size would need to be 1,802 people.