|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1)
Data - correlation and initial principal components-matrix |
|
Solution by SPSS |
|
|
(using the CanCorr macro) |
| 1.1 |
This is the
correlation-matrix of 8 items of an index of daily-life-activity of old-aged
people in a survey of rehabilitation after stroke |
|
|
|
|
|
Correlations for Set-1 |
|
correlations |
|
S1A455 |
S1A454 |
S1A461 |
S1A445 |
S1A446 |
S1A459 |
S1A448 |
S1A447 |
|
S1A455 S1A454 S1A461 S1A445 |
|
S1A455 |
Besuche erhalten |
1 |
0.585577 |
0.288664 |
0.280748 |
0.196817 |
0.198924 |
0.159755 |
0.219989 |
|
S1A455 1.0000
.5856 .2887 .2807 |
|
S1A454 |
Besuche machen
|
0.585577 |
1 |
0.268428 |
0.412443 |
0.248294 |
0.159625 |
0.219952 |
0.244413 |
|
S1A454 .5856
1.0000 .2684 .4124 |
|
S1A461 |
Briefe schreiben |
0.288664 |
0.268428 |
1 |
0.295551 |
0.249847 |
0.115058 |
0.173021 |
0.17995 |
|
S1A461 .2887
.2684 1.0000 .2956 |
|
S1A445 |
Spazieren |
0.280748 |
0.412443 |
0.295551 |
1 |
0.311627 |
0.223415 |
0.138686 |
0.279225 |
|
S1A445 .2807
.4124 .2956 1.0000 |
|
S1A446 |
Bücher lesen
|
0.196817 |
0.248294 |
0.249847 |
0.311627 |
1 |
0.357237 |
0.354167 |
0.358857 |
|
|
|
S1A459 |
Rätsellösen
|
0.198924 |
0.159625 |
0.115058 |
0.223415 |
0.357237 |
1 |
0.190039 |
0.267069 |
|
Correlations for Set-2 |
|
S1A448 |
Zeitschriften
|
0.159755 |
0.219952 |
0.173021 |
0.138686 |
0.354167 |
0.190039 |
1 |
0.387965 |
|
S1A446 S1A459 S1A448 S1A447 |
|
S1A447 |
Zeitungen
|
0.219989 |
0.244413 |
0.17995 |
0.279225 |
0.358857 |
0.267069 |
0.387965 |
1 |
|
S1A446 1.0000
.3572 .3542 .3589 |
|
S1A459 .3572
1.0000 .1900 .2671 |
| 1.2 |
This are the principal
components-loadings from that above correlations |
|
S1A448 .3542
.1900 1.0000 .3880 |
|
|
|
S1A447 .3589
.2671 .3880 1.0000 |
|
princ comps |
|
pc1 |
pc2 |
pc3 |
pc4 |
pc5 |
pc6 |
pc7 |
pc8 |
|
|
S1A455 |
Besuche erhalten |
0.6302 |
-0.5048 |
0.043 |
0.3458 |
-0.2521 |
0.0737 |
0.1294 |
-0.3753 |
|
Correlations Between Set-1
and Set-2 |
|
S1A454 |
Besuche machen
|
0.6845 |
-0.4792 |
0.0834 |
0.285 |
0.0603 |
-0.1517 |
-0.0102 |
0.4322 |
|
S1A446 S1A459 S1A448 S1A447 |
|
S1A461 |
Briefe schreiben |
0.5221 |
-0.2432 |
0.1443 |
-0.693 |
-0.3472 |
0.1998 |
-0.0407 |
0.0712 |
|
S1A455 .1968
.1989 .1598 .2200 |
|
S1A445 |
Spazieren |
0.6288 |
-0.2047 |
-0.2643 |
-0.2525 |
0.5731 |
-0.0929 |
-0.2514 |
-0.1696 |
|
S1A454 .2483
.1596 .2200 .2444 |
|
S1A446 |
Bücher lesen
|
0.6478 |
0.3853 |
-0.1337 |
-0.1657 |
-0.0437 |
-0.428 |
0.4482 |
-0.0229 |
|
S1A461 .2498
.1151 .1730 .1799 |
|
S1A459 |
Rätsellösen
|
0.4998 |
0.3554 |
-0.6417 |
0.1729 |
-0.3105 |
0.1516 |
-0.24 |
0.0725 |
|
S1A445 .3116
.2234 .1387 .2792 |
|
S1A448 |
Zeitschriften
|
0.5319 |
0.4527 |
0.5367 |
0.1119 |
-0.1095 |
-0.2134 |
-0.3847 |
-0.0777 |
|
|
S1A447 |
Zeitungen |
0.6155 |
0.3801 |
0.2087 |
0.1121 |
0.2928 |
0.5306 |
0.2258 |
0.0488 |
|
|
| 2)
Generating the canonical components by rotations of the loadingsmatrix |
|
|
|
| 2.1 |
First we collect the
variance of the first set of four variables (which explains also a part of
the variance of the second set). |
|
|
|
|
|
This variance should be in principal
components position (of that 4 items only). |
|
|
So we rotate for PC-position using
only the first four items as criterion for the routine |
|
|
Also we mark the found axes by
marker variables and call them pc1_x for them representing the pc's of the first set of items |
|
|
|
|
|
ax1 |
ax2 |
ax3 |
ax4 |
ax5 |
ax6 |
ax7 |
ax8 |
|
|
S1A455 |
Besuche erhalten |
0.7759 |
-0.3958 |
0.295 |
-0.3926 |
0 |
0 |
0 |
0 |
|
|
S1A454 |
Besuche machen
|
0.8219 |
-0.342 |
-0.031 |
0.4544 |
0 |
0 |
0 |
0 |
|
|
S1A461 |
Briefe schreiben |
0.5935 |
0.685 |
0.4181 |
0.0612 |
0 |
0 |
0 |
0 |
|
|
S1A445 |
Spazieren |
0.6748 |
0.2693 |
-0.6693 |
-0.1559 |
0 |
0 |
0 |
0 |
|
|
S1A446 |
Bücher lesen |
0.3431 |
0.1131 |
-0.0756 |
0.0059 |
0.5882 |
-0.3863 |
-0.4042 |
-0.453 |
|
|
S1A459 |
Rätsellösen |
0.242 |
0.0069 |
-0.0671 |
-0.0858 |
-0.2714 |
0.1811 |
-0.8347 |
-0.3555 |
|
|
S1A448 |
Zeitschriften |
0.2403 |
0.0214 |
0.0279 |
0.0674 |
-0.3731 |
-0.8486 |
-0.1238 |
-0.2487 |
|
|
S1A447 |
Zeitungen |
0.3198 |
0.0341 |
-0.0764 |
-0.0201 |
0.1010 |
-0.4877 |
-0.5889 |
0.5435 |
|
|
pc1_1 |
pc1_1 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
pc1_2 |
pc1_2 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
pc1_3 |
pc1_3 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
|
|
pc1_4 |
pc1_4 |
0 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
|
|
|
|
|
|
|
Obviously that variance needs only
4 dimensions/axes to be perfectly represented. And also obviously there is
partial variance of the items |
|
|
of the second set, which is not
explained by that first set and is shown in the axes 5-8 |
|
|
|
|
|
|
|
| 2.2 |
Next we collect the
variance of the second set of four variables (which explains also a part of
the variance of the first set). |
|
|
|
|
|
|
|
|
|
|
Again this variance should be in
principal components position (of that 4 items only). |
|
|
|
So we rotate for PC-position using
only the second four items as criterion for the routine |
|
|
|
Also we mark the new found axes by
marker variables and call them pc2_x for them representing the pc's of the second set of items |
|
|
|
|
|
|
|
|
|
|
ax1 |
ax2 |
ax3 |
ax4 |
ax5 |
ax6 |
ax7 |
ax8 |
|
|
S1A455 |
Besuche erhalten |
0.2763 |
0.0349 |
-0.0443 |
-0.0059 |
-0.7540 |
0.5614 |
-0.0760 |
0.1762 |
|
|
S1A454 |
Besuche machen |
0.3140 |
-0.0331 |
0.0020 |
-0.0330 |
-0.4078 |
0.1134 |
0.2115 |
0.8218 |
|
|
S1A461 |
Briefe schreiben |
0.2601 |
-0.0214 |
0.0541 |
-0.0709 |
-0.3204 |
-0.2465 |
-0.8157 |
0.3084 |
|
|
S1A445 |
Spazieren |
0.3427 |
0.0696 |
-0.0326 |
-0.1235 |
-0.6852 |
-0.584 |
0.2193 |
0.053 |
|
|
S1A446 |
Bücher lesen
|
0.7548 |
0.1466 |
0.4412 |
-0.4628 |
0 |
0 |
0 |
0 |
|
|
S1A459 |
Rätsellösen
|
0.6181 |
0.6996 |
-0.1296 |
0.3342 |
0 |
0 |
0 |
0 |
|
|
S1A448 |
Zeitschriften
|
0.6925 |
-0.5063 |
0.2544 |
0.4464 |
0 |
0 |
0 |
0 |
|
|
S1A447 |
Zeitungen |
0.7304 |
-0.2634 |
-0.5876 |
-0.2279 |
0 |
0 |
0 |
0 |
|
|
pc1_1 |
pc1_1 |
0.4115 |
0.0164 |
-0.0109 |
-0.0754 |
-0.7542 |
-0.0056 |
-0.1061 |
0.4944 |
|
|
pc1_2 |
pc1_2 |
0.0658 |
0.002 |
0.0554 |
-0.0837 |
0.0416 |
-0.72 |
-0.6647 |
-0.1537 |
|
|
pc1_3 |
pc1_3 |
-0.0687 |
-0.0622 |
0.044 |
0.0735 |
0.1615 |
0.633 |
-0.7272 |
0.1689 |
|
|
pc1_4 |
pc1_4 |
-0.0085 |
-0.1052 |
0.0687 |
0.0057 |
0.5092 |
-0.2391 |
0.1075 |
0.81 |
|
|
pc2_1 |
pc2_1 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
pc2_2 |
pc2_2 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
pc2_3 |
pc2_3 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
|
|
pc2_4 |
pc2_4 |
0 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
|
|
|
|
|
|
Again obviously that variance needs
only 4 dimensions/axes to be perfectly represented. And also obviously there
is partial variance of the items |
|
|
of the first set, which is not
explained by that second set and is now shown in the axes 5-8 |
|
|
|
|
|
|
| 2.3 |
From this we find the
canonical components of the second set and the crossloadings of the first set |
|
|
|
|
|
|
|
|
|
They are marked by the principal
components of the variance of the pcs of the first set, which is in common
with the variance of the second set |
|
|
So we rotate the subspace of first 4
axes of pc1_1 to pc1_4 into
principal components orientation |
|
|
|
and mark that axes by new variables cc2_1 to cc2_4 |
|
|
|
|
|
|
|
Cross Loadings for Set-1 |
|
|
|
ax1 |
ax2 |
ax3 |
ax4 |
ax5 |
ax6 |
ax7 |
ax8 |
|
1 2 3 4 |
|
S1A455 |
Besuche erhalten |
0.2727 |
0.0243 |
-0.0672 |
0.0072 |
-0.754 |
0.5614 |
-0.076 |
0.1762 |
|
S1A455 -.273 .024 .067 -.007 |
|
S1A454 |
Besuche machen |
0.3091 |
|
-0.0425 |
-0.0096 |
-0.4078 |
0.1134 |
0.2115 |
0.8218 |
|
S1A454 -.309
-.058 .043 .010 |
|
S1A461 |
Briefe schreiben |
0.2654 |
-0.069 |
0.0251 |
0.0146 |
-0.3204 |
-0.2465 |
-0.8157 |
0.3084 |
|
S1A461 -.265
-.069 -.025 -.015 |
|
S1A445 |
Spazieren |
0.3673 |
0.0515 |
0.0314 |
-0.0061 |
-0.6852 |
-0.584 |
0.2193 |
0.053 |
|
S1A445 -.367 .051 -.031 .006 |
|
S1A446 |
Bücher lesen |
0.8395 |
-0.1753 |
0.4428 |
0.2615 |
0 |
0 |
0 |
0 |
|
S1A446 -.840
-.175 -.443 -.261 |
|
S1A459 |
Rätsellösen |
0.5764 |
0.5187 |
-0.379 |
0.5051 |
0 |
0 |
0 |
0 |
|
S1A459 -.576 .519 .379 -.505 |
|
S1A448 |
Zeitschriften |
0.518 |
-0.6793 |
-0.5101 |
0.0999 |
0 |
0 |
0 |
0 |
|
S1A448 -.518
-.679 .510 -.100 |
|
S1A447 |
Zeitungen |
0.7571 |
0.0918 |
-0.2322 |
-0.6037 |
0 |
0 |
0 |
0 |
|
S1A447 -.757 .092 .232 .604 |
|
pc1_1 |
pc1_1
|
0.4177 |
-0.0166 |
-0.0245 |
0.0011 |
-0.7542 |
-0.0056 |
-0.1061 |
0.4944 |
|
Canonical Loadings for Set-2 |
|
pc1_2 |
pc1_2
|
0.0822 |
-0.0287 |
0.0819 |
0.0108 |
0.0416 |
-0.72 |
-0.6647 |
-0.1537 |
|
|
pc1_3 |
pc1_3
|
-0.09 |
-0.0765 |
-0.0408 |
0.0178 |
0.1615 |
0.633 |
-0.7272 |
0.1689 |
|
|
pc1_4 |
pc1_4 |
-0.0196 |
-0.1234 |
0.0096 |
-0.0137 |
0.5092 |
-0.2391 |
0.1075 |
0.81 |
|
|
pc2_1 |
pc2_1 |
0.9679 |
-0.1095 |
-0.2152 |
0.0701 |
0 |
0 |
0 |
0 |
|
|
|
pc2_2 |
pc2_2 |
0.0772 |
0.7851 |
0.1425 |
0.5977 |
0 |
0 |
0 |
0 |
|
|
|
pc2_3 |
pc2_3 |
-0.0279 |
-0.5975 |
0.4041 |
0.6921 |
0 |
0 |
0 |
0 |
|
|
pc2_4 |
pc2_4 |
-0.2377 |
-0.1208 |
-0.8775 |
0.3985 |
0 |
0 |
0 |
0 |
|
|
cc2_1 |
cc2_1 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
cc2_2 |
cc2_2 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
cc2_3 |
cc2_3 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
|
|
cc2_4 |
cc2_4 |
0 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
|
|
|
|
|
|
|
The loadings of the items-sets in
the first four axes are the "cross-" resp. the "canonical
loadings" |
|
|
|
|
|
|
|
| 2.4 |
Next we find the
canonical components of the first set |
|
|
|
|
|
|
|
|
|
|
They are marked by the principal
components of the variance of the pcs of the second set, which is in common
with the variance of the first set |
|
|
We have to collect the variance of
the first set into the first four axes. |
|
|
|
|
After that we rotate the subspace
of first 4 axes into principal components orientation for the
princ.components of the second set pc2_1 to pc2_4 |
|
|
and mark that axes by new variables cc1_1 to cc1_4 |
|
|
|
|
|
|
|
Canonical Loadings for Set-1 |
|
|
|
ax1 |
ax2 |
ax3 |
ax4 |
ax5 |
ax6 |
ax7 |
ax8 |
|
1 2 3 4 |
|
S1A455 |
Besuche erhalten |
0.6261 |
|
-0.7059 |
-0.2881 |
0 |
0 |
0 |
0 |
|
S1A455 -.626 .163 .706 -.288 |
|
S1A454 |
Besuche machen |
0.7097 |
-0.3865 |
-0.4464 |
0.3843 |
0 |
0 |
0 |
0 |
|
S1A454 -.710
-.387 .446 .384 |
|
S1A461 |
Briefe schreiben |
0.6093 |
-0.4636 |
0.2636 |
-0.5868 |
0 |
0 |
0 |
0 |
|
S1A461 -.609
-.464 -.264 -.587 |
|
S1A445 |
Spazieren |
0.8432 |
0.3458 |
0.3296 |
0.2467 |
0 |
0 |
0 |
0 |
|
S1A445 -.843 .346 -.330 .247 |
|
S1A446 |
Bücher lesen |
0.3657 |
-0.0261 |
0.0422 |
-0.0065 |
0.7117 |
0.4978 |
-0.0547 |
-0.3262 |
|
S1A446 -.366
-.026 -.042 -.007 |
|
S1A459 |
Rätsellösen |
0.2511 |
|
-0.0361 |
-0.0126 |
0.6191 |
-0.6138 |
0.0089 |
-0.4116 |
|
S1A459 -.251 .077 .036 -.013 |
|
S1A448 |
Zeitschriften |
0.2257 |
-0.1011 |
-0.0486 |
-0.0025 |
0.3207 |
0.1475 |
0.8993 |
-0.0568 |
|
S1A448 -.226
-.101 .049 -.002 |
|
S1A447 |
Zeitungen |
0.3298 |
0.0137 |
-0.0221 |
0.015 |
0.6737 |
-0.0451 |
0.1566 |
0.6403 |
|
S1A447 -.330 .014 .022 .015 |
|
pc1_1 |
pc1_1 |
0.959 |
-0.1116 |
-0.257 |
-0.0429 |
0 |
0 |
0 |
0 |
|
Cross Loadings for Set-2 |
|
pc1_2 |
pc1_2 |
0.1887 |
-0.1926 |
0.8602 |
-0.4329 |
0 |
0 |
0 |
0 |
|
|
pc1_3 |
pc1_3 |
-0.2066 |
-0.5136 |
-0.4289 |
-0.7139 |
0 |
0 |
0 |
0 |
|
|
pc1_4 |
pc1_4 |
-0.0449 |
-0.8287 |
0.1005 |
0.5488 |
0 |
0 |
0 |
0 |
|
|
pc2_1 |
pc2_1
|
0.4216 |
-0.0163 |
-0.0205 |
-0.0017 |
0.8316 |
0.0334 |
0.357 |
-0.0368 |
|
|
pc2_2 |
pc2_2
|
0.0336 |
0.1169 |
0.0136 |
-0.0149 |
0.2362 |
-0.5011 |
-0.5956 |
-0.5685 |
|
|
pc2_3 |
pc2_3
|
-0.0121 |
-0.0889 |
0.0385 |
-0.0172 |
-0.1294 |
0.5844 |
0.1794 |
-0.7744 |
|
|
pc2_4 |
pc2_4 |
-0.1035 |
-0.018 |
-0.0836 |
-0.0099 |
-0.2302 |
-0.6227 |
0.6828 |
-0.2736 |
|
Canonical Correlations |
|
cc2_1 |
cc2_1 |
0.4356 |
0 |
0 |
0 |
0.8815 |
0.1254 |
0.1323 |
0.0071 |
|
1
.436 |
|
cc2_2 |
cc2_2 |
0 |
0.1489 |
0 |
0 |
0.1995 |
-0.671 |
-0.6963 |
0.0534 |
|
2
.149 |
|
cc2_3 |
cc2_3 |
0 |
|
0.0952 |
0 |
0.0044 |
0.7041 |
-0.6884 |
-0.1459 |
|
3
.095 |
|
cc2_4 |
cc2_4 |
0 |
0 |
0 |
-0.0249 |
0.0182 |
-0.1409 |
0.0653 |
-0.9874 |
|
4
.025 |
|
cc1_1 |
cc1_1 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
cc1_2 |
cc1_2 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
0 |
|
|
cc1_3 |
cc1_3 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
0 |
|
|
cc1_4 |
cc1_4 |
0 |
0 |
0 |
1 |
0 |
0 |
0 |
0 |
|
|
|
The loadings of the items-sets in
the first four axes are the "cross-" resp. the "canonical
loadings" |
|
|
|
After this we have all needed
coefficients at hand; the canonical correlations can directly be read off
from the correlations between the canonical |
|
|
components of the two sets |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 3)
Regression: computing the canonical coefficients as beta-weights |
|
|
|
|
The "canonical coefficients" are the beta-weights of the canonical
components in terms of the set-variables. |
|
|
|
|
|
In the above table 2.4 we
postmultiply the four leading columns by the inverse of the loadings of the
first set-items |
|
|
and get the following beta-weights
table |
|
|
| 3.1 |
|
beta-weights |
|
S1A455 |
S1A454 |
S1A461 |
S1A445 |
|
|
|
|
|
|
S1A455 |
Besuche erhalten |
1 |
0 |
0 |
0 |
|
|
|
|
S1A454 |
Besuche machen
|
0 |
1 |
0 |
0 |
|
|
|
|
S1A461 |
Briefe schreiben |
0 |
0 |
1 |
0 |
|
|
|
|
S1A445 |
Spazieren |
0 |
0 |
0 |
1 |
|
|
|
|
|
|
S1A446 |
Bücher lesen |
0.035447 |
0.097918 |
0.149133 |
0.217214 |
|
|
|
|
S1A459 |
Rätsellösen |
0.145499 |
-0.004887 |
0.021713 |
0.178165 |
|
|
|
|
S1A448 |
Zeitschriften |
0.022548 |
0.163333 |
0.113364 |
0.031485 |
|
|
|
|
S1A447 |
Zeitungen |
0.09104 |
0.091576 |
0.071528 |
0.194756 |
|
|
|
|
|
|
pc1_1 |
pc1_1 |
0.37212 |
0.39418 |
0.284613 |
0.323593 |
|
|
|
|
pc1_2 |
pc1_2 |
-0.485467 |
-0.419401 |
0.840131 |
0.330226 |
|
|
|
|
pc1_3 |
pc1_3 |
0.415143 |
-0.043607 |
0.588262 |
-0.94168 |
|
|
|
|
pc1_4 |
pc1_4 |
-1.010086 |
1.16902 |
0.157539 |
-0.401022 |
|
|
|
|
|
|
pc2_1 |
pc2_1 |
0.101179 |
0.127692 |
0.130669 |
0.222987 |
|
|
|
|
pc2_2 |
pc2_2 |
0.085559 |
-0.114602 |
-0.046841 |
0.10665 |
|
|
|
|
pc2_3 |
pc2_3 |
-0.082015 |
0.050822 |
0.08014 |
-0.054175 |
|
|
|
|
pc2_4 |
pc2_4 |
0.037329 |
0.008835 |
-0.047569 |
-0.123561 |
|
|
|
|
|
|
cc2_1 |
cc2_1 |
0.097946 |
0.111228 |
0.131929 |
0.254932 |
|
|
|
|
cc2_2 |
cc2_2 |
0.100593 |
-0.13539 |
-0.093217 |
0.106617 |
|
|
|
|
cc2_3 |
cc2_3 |
-0.075483 |
-0.031027 |
0.039331 |
0.053744 |
|
|
|
Raw Canonical Coefficients
for Set-1 |
|
cc2_4 |
cc2_4 |
0.016352 |
-0.020855 |
0.017667 |
-0.007355 |
|
|
|
|
|
1 2 3 4 |
|
cc1_1 |
cc1_1 |
0.224853 |
|
0.302866 |
0.585241 |
|
|
|
S1A455 -.225 .675 .792 -.656 |
|
cc1_2 |
cc1_2 |
0.675776 |
-0.909544 |
-0.626228 |
0.716246 |
|
|
|
S1A454 -.255
-.909 .326 .837 |
|
cc1_3 |
cc1_3 |
-0.79278 |
-0.325869 |
0.413078 |
0.564461 |
|
|
|
S1A461 -.303
-.626 -.413 -.709 |
|
cc1_4 |
cc1_4 |
-0.656534 |
0.837331 |
-0.709346 |
0.295316 |
|
|
|
|
|
S1A445 -.585 .716
-.564 .295 |
|
|
After rerotating of matrix 2.4 to
collect the variance of the second item-set in the first four axes we
postmultiply |
|
(note: these are displayed in transposition) |
|
the four leading columns by the inverse of the loadings of the
second set-items |
|
|
| 3.2 |
|
beta-weights |
|
S1A446 |
S1A459 |
S1A448 |
S1A447 |
|
|
|
|
|
|
S1A455 |
Besuche erhalten |
0.085513 |
0.121884 |
0.053553 |
0.135974 |
|
|
|
|
S1A454 |
Besuche machen |
0.142712 |
0.051567 |
0.105943 |
0.138325 |
|
|
|
|
S1A461 |
Briefe schreiben |
0.191454 |
0.011566 |
0.071874 |
0.080271 |
|
|
|
|
S1A445 |
Spazieren |
0.219747 |
0.101228 |
-0.030166 |
0.185036 |
|
|
|
|
|
|
S1A446 |
Bücher lesen
|
1 |
0 |
0 |
0 |
|
|
|
|
S1A459 |
Rätsellösen
|
0 |
1 |
0 |
0 |
|
|
|
|
S1A448 |
Zeitschriften
|
0 |
0 |
1 |
0 |
|
|
|
|
S1A447 |
Zeitungen |
0 |
0 |
0 |
1 |
|
|
|
|
|
|
pc1_1 |
pc1_1 |
0.213674 |
0.101731 |
0.072384 |
0.187846 |
|
|
|
|
pc1_2 |
pc1_2 |
0.132045 |
-0.037653 |
-0.020009 |
0.004518 |
|
|
|
|
pc1_3 |
pc1_3 |
-0.06503 |
-0.04017 |
0.0883 |
-0.076607 |
|
|
|
|
pc1_4 |
pc1_4 |
0.022495 |
-0.101603 |
0.093176 |
-0.037198 |
|
|
|
|
|
|
pc2_1 |
pc2_1 |
0.384159 |
0.314588 |
0.352483 |
0.371729 |
|
|
|
|
pc2_2 |
pc2_2 |
0.175166 |
0.836139 |
-0.605124 |
-0.314839 |
|
|
|
|
pc2_3 |
pc2_3 |
0.710012 |
-0.208483 |
0.409397 |
-0.945518 |
|
|
|
Raw Canonical Coefficients
for Set-2 |
|
pc2_4 |
pc2_4 |
-0.801914 |
0.579145 |
0.773585 |
-0.394924 |
|
|
|
|
|
1 2 3 4 |
|
cc2_1 |
cc2_1 |
0.556148 |
0.237188 |
0.09919 |
0.455684 |
|
|
|
S1A446 -.556
-.232 -.932 -.303 |
|
cc2_2 |
cc2_2 |
-0.231875 |
|
-0.851739 |
0.324732 |
|
|
|
S1A459 -.237 .676 .541 -.608 |
|
cc2_3 |
cc2_3 |
0.932906 |
-0.541039 |
-0.67551 |
-0.160372 |
|
|
|
S1A448 -.099
-.851 .675 -.255 |
|
cc2_4 |
cc2_4 |
0.303422 |
0.60837 |
0.254649 |
-0.973888 |
|
|
|
|
|
S1A447 -.455 .325 .160 .973 |
|
cc1_1 |
cc1_1 |
0.242259 |
0.103319 |
0.043207 |
0.198496 |
|
|
|
|
cc1_2 |
cc1_2 |
-0.034516 |
0.100725 |
-0.126786 |
0.048338 |
|
|
|
(note: these are displayed in transposition) |
|
cc1_3 |
cc1_3 |
0.088825 |
-0.051514 |
-0.064318 |
-0.01527 |
|
|
|
|
cc1_4 |
cc1_4 |
-0.007557 |
-0.015152 |
-0.006342 |
0.024256 |
|
|
|
|
|
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|