1. Fill-in-the-blanks:

    In a one-way ANOVA, SStotal is partitioned into 2 components:

    (1) _______________ (variance due to the treatment effect) and (2) _______________ (unexplained variance).

    We divide SS by their df to get ____________ which represent ________________________________________.

    We then calculate the ratio of those MS values and compare it to the _______________ distribution.

    Under the null hypothesis, we expect this ratio to equal _______________.

    If the null hypothesis is false, this ratio will be _________________________.

    The power of a statistical test refers to ____________________________________________________________.

    If we want to increase the power of our ANOVA, we could:

      (a) ______________________________ our sample size,

      (b) ______________________________ our alpha-level, or

      (c) ______________________________ the size of MSE.

    If MSE is small, our MSR will be ____________________ and we have a better chance of rejecting our null hypothesis.

  1. The following figures attempt to demonstrate how the total variation is partitioned under one-way, AxB, and AxS (repeated measures) ANOVA. Explain how the AxB and AxS work to increase the power of our test.