Factor Analysis of Duchenne Muscular Dystrophy Data 2

Because we are going to extract five components the communalities are 1.0. Factor scores have been saved.

Communalities
  Initial Extraction
age 1.000 1.000
creatine kinase 1.000 1.000
hemopexin 1.000 1.000
lactate dehydrogenase 1.000 1.000
pyruvate kinase 1.000 1.000
Extraction Method: Principal Component Analysis.

Five components are extracted and this time the axes are rotated. The initial eigen values and Extraction Sums of Squared Loadings are identical to the previous analysis. However, because we have now asked for axis rotation we are provided with additional information about the Rotation Sum of Squared loadings.

Total Variance Explained
  Initial Eigenvalues Extraction Sums of Squared Loadings Rotation Sums of Squared Loadings
Component Total % Variance Cumulative % Total % Variance Cumulative % Total % Variance Cumulative %
1 2.576 51.515 51.515 2.576 51.515 51.515 1.705 34.100 34.100
2 1.189 23.779 75.295 1.189 23.779 75.295 1.023 20.454 54.553
3 .630 12.594 87.889 .630 12.594 87.889 1.014 20.280 74.834
4 .445 8.909 96.798 .445 8.909 96.798 1.013 20.262 95.096
5 .160 3.202 100.000 .160 3.202 100.000 .245 4.904 100.000
Extraction Method: Principal Component Analysis.

Examining the rotated factor loadings (below) it is apparent that the first component is almost entirely creatine kinase and lactate dehydrogenase (plus some pyruvate kinase). The second component is almost entirely pyruvate kinase. If we wished to look further factor three is an age factor and factor four is hemopexin.

Rotated Component Matrix(a)

Component
1 2 3 4 5
age 0.053 0.105 0.977 0.176 -0.003
creatine kinase 0.965 0.218 0.090 0.087 -0.070
hemopexin 0.093 0.153 0.182 0.967 0.022
lactate dehydrogenase 0.806 0.328 -0.018 0.086 0.484
pyruvate kinase 0.335 0.913 0.131 0.181 0.071
Extraction Method: Principal Component Analysis.
Rotation Method: Varimax with Kaiser Normalization.
a Rotation converged in 6 iterations.

Again we have plotted the factor scores and overlaid the plot with information about carrier status.

Factor 1 v factor 2 scores scatter plot

This time the plot shows that there is a lot of overlap between the two groups. As before the carriers are more heterogeneous, especially on factor 1. What the analysis tells us is that carriers are not characterised by one set of enzyme values, it may be that carriers are composed of different types.

Back to PCA Example 2 menu.

Valid HTML 4.01!