Two-population tests for differences between proportions follow the same logic as those for differences between means. The IV is still a two-class, categorical measure; however, the DV is a proportion rather than a mean. Differences between proportions have their own sampling distribution, which looks very much like that for differences between means and can be drawn.
Population proportions are symbolized with the letter P and sample proportions with (pronounced “p hat”). Like confidence intervals for proportions, two-population tests employ the z distribution. Thus, the sample size must be at least 100; these tests cannot be carried out on samples smaller than that. All the other fundamentals for proportion tests are the same as those for tests of differences between means, so we will dive right into an example.