Tests of Differences between Proportions (5 of 5)

previous


Summary of Computations
  1. Compute p1 and p2.

  2. Compute

  3. Compute .

  4. Compute .

  5. Use a z table to compute the probability value from z. Note that the correction for continuity is not used in the test for differences between proportions.

Assumptions
  1. 1. The two proportions are independent.

  2. For the normal approximation to be adequate, π should not be too close to 0 or to 1. Values between 0.10 and 0.90 allow the approximation to be adequate.

  3. For the normal approximation to be adequate, there should be at least 10 subjects per group.



previous