Capítulo 5: diferencias que presenta una variable

Multiple/Post Hoc Group Comparisons in ANOVA
Note: We may just go over this quickly in class. The key thing to understand is that, when trying to identify where differences are between groups, there are different ways of adjusting the probability estimates to reflect the fact that multiple comparisons are being made. Introduction. In a one-way ANOVA, the F statistic tests whether the treatmenteffects are all equal, i.e. that there are no differences among the means of the J groups. A significant F value indicates that there are differences in the means, but it does not tell you where those differences are, e.g. group 1’s mean might be different than group 2’s mean but not different from group 3’s mean. To isolate where the differences are, you could do a series of pairwise T-tests. Theproblem with this is that the significance levels can be misleading. For example, if you have 7 groups, there will be 21 pairwise comparisons of means; if using the .05 level of significance, you would expect at least one statistically significant difference even if no differences exist. Therefore, various methods have been developed for doing multiple comparisons of group means. In SPSS, one wayto accomplish this is via the use of the /POSTHOC parameter on the Oneway command. We’ll present the SPSS output and then explain what the different parts mean.
ONEWAY score BY program /STATISTICS DESCRIPTIVES /MISSING ANALYSIS /POSTHOC = LSD BONFERRONI SIDAK SCHEFFE ALPHA(.05).

Oneway
Descriptives SCORE 95% Confidence Interval for Mean Lower Bound Upper Bound 9.4116 14.1884 6.7597 10.840310.5811 13.8189 6.7169 10.4831 9.2950 11.4050

N 1.00 2.00 3.00 4.00 Total 5 5 5 5 20

Mean 11.8000 8.8000 12.2000 8.6000 10.3500

Std. Deviation 1.92354 1.64317 1.30384 1.51658 2.25424

Std. Error .86023 .73485 .58310 .67823 .50406

Minimum 9.00 6.00 11.00 7.00 6.00

Maximum 14.00 10.00 14.00 11.00 14.00

ANOVA
SCORE
Sum of Squares 54.950 41.600 96.550
df
Mean Square 18.317 2.600F 7.045
Sig. .003

Between Groups Within Groups Total

3 16 19

Multiple/Post Hoc Group Comparisons in Anova - Page 1

Post Hoc Tests
Multiple Comparisons Dependent Variable: SCORE Mean Difference Std. Error (I-J) 3.0000* 1.01980 -.4000 1.01980 3.2000* 1.01980 -3.0000* 1.01980 -3.4000* 1.01980 .2000 1.01980 .4000 1.01980 3.4000* 1.01980 3.6000* 1.01980 -3.2000* 1.01980 -.2000 1.01980-3.6000* 1.01980 3.0000 1.01980 -.4000 1.01980 3.2000* 1.01980 -3.0000 1.01980 -3.4000* 1.01980 .2000 1.01980 .4000 1.01980 3.4000* 1.01980 3.6000* 1.01980 -3.2000* 1.01980 -.2000 1.01980 -3.6000* 1.01980 3.0000 1.01980 -.4000 1.01980 3.2000* 1.01980 -3.0000 1.01980 -3.4000* 1.01980 .2000 1.01980 .4000 1.01980 3.4000* 1.01980 3.6000* 1.01980 -3.2000* 1.01980 -.2000 1.01980 -3.6000* 1.01980 3.00001.01980 -.4000 1.01980 3.2000* 1.01980 -3.0000 1.01980 -3.4000* 1.01980 .2000 1.01980 .4000 1.01980 3.4000* 1.01980 3.6000* 1.01980 -3.2000* 1.01980 -.2000 1.01980 -3.6000* 1.01980

LSD

Bonferroni

Sidak

Scheffe

(I) PROGRAM (J) PROGRAM 1.00 2.00 3.00 4.00 2.00 1.00 3.00 4.00 3.00 1.00 2.00 4.00 4.00 1.00 2.00 3.00 1.00 2.00 3.00 4.00 2.00 1.00 3.00 4.00 3.00 1.00 2.00 4.00 4.00 1.002.00 3.00 1.00 2.00 3.00 4.00 2.00 1.00 3.00 4.00 3.00 1.00 2.00 4.00 4.00 1.00 2.00 3.00 1.00 2.00 3.00 4.00 2.00 1.00 3.00 4.00 3.00 1.00 2.00 4.00 4.00 1.00 2.00 3.00

Sig. .009574 .700062 .006355 .009574 .004207 .846988 .700062 .004207 .002781 .006355 .846988 .002781 .057442 1.000000 .038130 .057442 .025242 1.000000 1.000000 .025242 .016686 .038130 1.000000 .016686 .056084 .999272 .037530.056084 .024978 .999987 .999272 .024978 .016571 .037530 .999987 .016571 .068155 .984100 .048181 .068155 .033774 .997930 .984100 .033774 .023519 .048181 .997930 .023519

95% Confidence Interval Lower Bound Upper Bound .8381 5.1619 -2.5619 1.7619 1.0381 5.3619 -5.1619 -.8381 -5.5619 -1.2381 -1.9619 2.3619 -1.7619 2.5619 1.2381 5.5619 1.4381 5.7619 -5.3619 -1.0381 -2.3619 1.9619 -5.7619 -1.4381...