Home page for accesible maths 5 Analysis of Variance

Style control - access keys in brackets

Font (2 3) - + Letter spacing (4 5) - + Word spacing (6 7) - + Line spacing (8 9) - +

5.3 Summary

{mdframed}
  • A hypothesis tests can result in one of two errors:

    • Reject the null hypothesis when it is in fact true. This is a Type I error.

    • Accept the null hypothesis when the alternative hypothesis is in fact true. This is a Type II error.

  • The probability of making a Type I error is equal to the significance level of the test, by the very definition of the significcance level.

  • The power of the test is the probability of correctly accepting the alternative hypothesis. This probability is one minus the probability of making a Type II error.

  • In multiple testing, for example when many groups are compared in a pairwise manner, the probability of making a Type I error in at least one of these tests becomes magnified.

  • The family wide error rate is defined as the probability of incorrectly rejecting at least one of the null hyopothesis. In multiple testing, a value for the FWER is usually specified. This is then used to imply the signficance level for the individual tests.

  • An alternative to mutiple two sample t-tests when we want to compare means across three or more groups is the ANOVA.

  • Given k groups, ANOVA is used to test

    H0:μ1=μ2==μk

    vs.

    H1:μ1μ2μk.