Table 3.
Non-Euclidean multivariate analysis of variance
| Factor | df | Sum of squares | Pseudo-F-statistic | P value a |
|---|---|---|---|---|
| Body siteb |
3 |
1.363 |
1.624 |
0.00015 |
| Cutaneotype |
1 |
0.626 |
2.236 |
0.00015 |
| SampleMonthYear |
24 |
10.141 |
1.510 |
0.00015 |
| Status |
2 |
0.854 |
1.526 |
0.0003 |
| Skin micro-environment |
1 |
0.321 |
1.141 |
0.11000 |
| Gender |
1 |
0.363 |
1.296 |
0.2940 |
| Ethnicity |
3 |
0.972 |
1.158 |
0.3179 |
| Age |
1 |
0.305 |
1.091 |
1.0 |
| History |
1 |
0.302 |
1.080 |
1.0 |
| Status × body site |
6 |
1.529 |
0.910 |
1.0 |
| Status × age |
2 |
0.482 |
0.862 |
1.0 |
| Status × sample month year |
28 |
7.265 |
0.927 |
1.0 |
| Status × gender |
2 |
0.468 |
0.837 |
1.0 |
| Status × ethnicity |
6 |
1.733 |
1.032 |
1.0 |
| Status × history |
2 |
0.490 |
0.876 |
1.0 |
| Status × cutaneotype |
2 |
0.549 |
0.981 |
1.0 |
| Residuals |
66 |
18.471 |
|
|
| Total | 150 | 45.913 |
aP values are based on 100,000 permutations and are adjusted for multiple testing using the Bonferroni method. Significant values are in bold.
bBody sites are grouped into head, body, upper extremities, lower extremities.