LESSON 9

Dr. John M. Ritz

t-Test

    * A t-test is used to determine if there is a significant difference between two sample means.
    * Typically, one mean represents an experimental group receiving some treatment and the other mean represents a control group.
    * H1 : Students receiving individualized instruction will score higher when evaluated than those exposed to the group method of instruction.
    * Control Group - 21, 20, 18, 15, 15, 12, 10 and 6.
      Experimental Group - 25, 24, 22, 22, 21, 21, 18 and 17

t-Test Procedures

    * Determine the sum of both sets of scores.
    * Find the mean of both sets of scores.
    * Find the difference from the mean of both sets of scores.
    * Square the difference of both groups.
    * Find the sums of the squared differences of both groups.
    * Substitute into the formula and solve.

Chi-Square

    * Chi-square is a means of answering questions about data existing in the form of "frequencies", rather than as scores or measurements along some scale.
    * Nominal data is used.
    * Typically, the question we want answered when we have such frequency data is whether the frequencies observed in our sample deviate significantly from some theoretical or expected populations frequencies.

Chi-Square

    * Continued . . . Examples of this kind include male/female, pass/fail, achiever/non-achiever, soda/fruit drinks, mountains/beaches, etc.
    * H0 : There is no significant genetic correlation when comparing peoples eye and hair color.

Chi-square Procedures

    * Set up matrix.
    * Sum A + B
    * Sum C + D
    * Sum A + C
    * Sum B + D
    * Substitute into the formula and solve