Chapter 13 Problems

2. Both the F-ratio and the t statistic are comparing the actual mean difference between samples (numerator) with the difference that would be expected from chance factors such as experimental error (denominator). If the numerator is sufficiently bigger than the denominator, we conclude that there is a significant difference between treatments.

4. With 3 or more treatment conditions you need 2 or more t tests to evaluate all of the mean differences. Each test involves a risk of a Type I error. The more tests you do the more risk there is of a Type I error. The ANOVA performs all of the tests simultaneously with a single, fixed alpha level.

6. a. There is n difference between treatments: Both have T = 8 and Mean = 2. MS-between should be zero.

b. F = 0

c. SS-between = 0 and MS-between = 0

8. a. The null hypothesis states that there are no differences among the treatments, Ho: mu-1, = mu-2 = mu-3. The critical value for alpha is 4.26.

     Source             SS      df       MS
Between Treatments      56       2       28   F(2,9) = 7.0
Within Treatments       36       9        4
Total                   92      11

Reject Ho and conclude that there are significant differences among the three treatments.

b. Tukey's HSD = 3.95. Treatments I and II are both significantly different from treatment III, but there is no significant difference between I and II.

10. a. The null hypothesis states that the size of discrepancy has no effect on attitude change, Ho: mu-1 = mu-2 = mu-3. The critical value of alpha (.01) is 6.36.

    Source            SS      df       MS
Between Treatments    36       2       18   F(2,15) =  9.0
Within Treatments     30      15        4
Total                 66      17

Reject Ho

b. The greatest change in attitude occurs when there is a moderate discrepancy between the persuasive argument and a person's original opinion.

12. a. The null hypothesis states that handedness has no effect on discrimination, Ho: mu-1 = mu-2 = mu-3. The critical value of alpha (.05) is 3.88.

    Source            SS      df       MS
Between Treatments    30       2       15   F(2,12) = 15.00
Within Treatments     12      12        1
Total                 42      14

Reject Ho and conclude that there are significant differences among the three groups.

b. Of the three samples the largest variance is for the righ-handed subjects (variance = 1.5) and the smallest is for the left-handed subjects (variance = 0.50). F-max = 3.00 which is not significant. The homogeneity assumption is satisfied.

14.

    Source            SS      df       MS
Between Treatments    48       2       24   F(2,27) = 12.00
Within Treatments     54      27        2
Total                102      29

16.

    Source            SS      df       MS
Between Treatments    30       3       10   F(3,20) = 5.00
Within Treatments     40      20        2
Total                 70      23

18. The null hypothesis states that there are no differences among the three treatments, Ho: mu-1 = mu-2 = mu-3. The critical value for alpha (.05) is 3.88.

    Source            SS      df       MS
Between Treatments    15       2       7.5   F(2,12) = 1.50
Within Treatments     60      12        5
Total                 75      14

Fail to reject Ho. No evidence of any significant differences.

20. a. The means and SS values are:

           Endormorph     Ectomorph      Mesomorph
          Mean = 22.0    Mean = 17.6    Mean = 15.0
            SS = 24.0      SS = 23.2      SS = 30.0

The null hypothesis states that there are no differences in personality among the three body types, Ho: mu-1 = mu-2 = mu-3. The critical value for alpha (.05) is 3.35. The analysis of variance produces,

    Source             SS     df       MS
Between Treatments   125.2    2       62.6   F(2,12) = 9.74
Within Treatments     77.2   12        6.43
Total                202.4   14

Reject Ho and conclude that there significant differences in language development among the three groups.

22. The means and SS values are

           Alphas           Betas          Gammas
          Mean = 44.2    Mean = 47.1    Mean = 36.8
          SS = 865.6     SS = 500.9      SS = 325.6

The null hypothesis states that there are no differences in health among the three groups, Ho: mu-1 = mu-2 = mu-3. The critical value for alpha (.05) is 3.35. The analysis of variances produces,

    Source            SS      df       MS
Between Treatments   564.2    2      282.1   F(2,27) = 9.74
Within Treatments   1692.1   27       62.67
Total               2256.3   29

Reject Ho. There are significant differences among the three groups.

b. Beginning with the largest mean difference, the Scheffe test results are:

Betas versus Gammas: F(2,27) = 4.23 (significant). Alphas versus Gammas: F(2,27) = 2.18 (not significant) The only significant difference is between the Betas and the Gammas. The Alphas are an intermediate group, not significantly different from either extreme.