Abstract
This paper sets out to fill a gap by providing a quick method, with few restrictions, for testing whether the difference between a pair of proportions is significant. Alternative methods are noted and their limitations mentioned briefly. The technique presented here depends on a set of tables based on Zubin's t. It may be applied to any nominal data that can be arranged in 2 × 2 contingency tables. There is an illustration of how the method may be used including a caution about its indiscriminate application.