PPT Slide

assumptions for tests with z-scores

• using the z-score as test statistics is only one of many possible ways to test a hypothesis, all requiring certain conditions

• the assumptions for validity of z-score tests are:
  • random sampling (sample must be representative)
  • independent observations (no predictable relationships between individual events)
  • the value of s is unchanged by treatment (standard error of the mean for z-score requires s to be identical before/after treatment: only shift of mean)
  • normal sampling distribution (basis for applying unit normal table)

Previous slide

Next slide

Back to first slide

View graphic version