Correlations Between Cooijmans Tests

General remarks

In these analyses, each test's loading is given on the first principal factor - g - in a matrix of tests for which all intercorrelations are known. It takes years to find out all those correlations, so on this page one will find mainly the older tests.

A test's g factor loading is an indicator of validity, of how well the test succeeds in measuring intelligence. To my experience, it is not a simple case of "higher is better"; test quality is also determined by other factors like robustness and reliability, which are partly independent of validity. The best tests have loadings between .7 and .9. Below .7 the test apparently does not catch in a wide enough range of abilities, for instance because it has too few items. Above .9 the test does catch in a very wide range of abilities, but that is no guarantee that its robustness is in order. The Final Test for instance comes out on top in these analyses, but is clearly not the best test as it is also the one that has suffered the most from raw score inflation over the years; it lacks robustness.

Abbreviations used

an - Analogies of Long Test For Genius
an1 - Analogies #1
ass - Association subtest of Long Test For Genius
cit - Cooijmans Intelligence Test Form 1
fin - The Final Test
gat - Genius Association Test
ltfg - Long Test For Genius
num - Numbers
qmc3 - Qoymans Multiple-Choice #3
stfg - Short Test For Genius
sth - Space, Time and Hyperspace
tteat - The Test To End All Tests

Comment to analyses of August 2004

Three sets of tests are analyzed to avoid the inclusion of overlapping tests (containing same items) in each particular analysis. For each analysis, given are the tests' loadings on the first principal factor - g - , the percentage of the variance in the correlation matrix accounted for by this factor, and the matrix itself.

Analysis #1 - August 2004

fin 0.92
cit 0.92
tteat 0.85
ltfg 0.84
num 0.81
qmc3 0.66
an1 0.65

This factor accounts for 66 % of the variance in this matrix.

Test	ltfg	fin	tteat	cit	num	qmc3	an1
ltfg	1	0.86	0.73	0.72	0.57	0.57	0.32
fin	0.86	1	0.8	0.69	0.71	0.55	0.62
tteat	0.73	0.8	1	0.8	0.85	0.36	0.27
cit	0.72	0.69	0.8	1	0.41	0.77	0.82
num	0.57	0.71	0.85	0.41	1	0.47	0.59
qmc3	0.57	0.55	0.36	0.77	0.47	1	0.03
an1	0.32	0.62	0.27	0.82	0.59	0.03	1

Analysis #2 - August 2004

cit 0.95
fin 0.94
tteat 0.89
an1 0.76
stfg 0.7
qmc3 0.45

This factor accounts for 64 % of the variance in this matrix.

Test	stfg	fin	tteat	cit	qmc3	an1
stfg	1	0.77	0.93	0.37	-0.61	0.82
fin	0.77	1	0.8	0.69	0.55	0.62
tteat	0.93	0.8	1	0.8	0.36	0.27
cit	0.37	0.69	0.8	1	0.77	0.82
qmc3	-0.61	0.55	0.36	0.77	1	0.03
an1	0.82	0.62	0.27	0.82	0.03	1

Analysis #3 - August 2004

cit 0.94
fin 0.94
tteat 0.79
an 0.79
num 0.79
sth 0.76
ass 0.76
an1 0.67
qmc3 0.6

This factor accounts for 62 % of the variance in this matrix.

Test	sth	fin	tteat	cit	num	qmc3	an1	an	ass
sth	1	0.71	0.51	0.63	0.57	0.35	0.56	0.53	0.49
fin	0.71	1	0.8	0.69	0.71	0.55	0.62	0.78	0.75
tteat	0.51	0.8	1	0.8	0.85	0.36	0.27	0.59	0.39
cit	0.63	0.69	0.8	1	0.41	0.77	0.82	0.73	0.77
num	0.57	0.71	0.85	0.41	1	0.47	0.59	0.38	0.55
qmc3	0.35	0.55	0.36	0.77	0.47	1	0.03	0.44	0.24
an1	0.56	0.62	0.27	0.82	0.59	0.03	1	0.42	0.43
an	0.53	0.78	0.59	0.73	0.38	0.44	0.42	1	0.7
ass	0.49	0.75	0.39	0.77	0.55	0.24	0.43	0.7	1

Analysis #4 - Verbal tests - March 2006

fin 0.95
an 0.87
gat 0.81
qmc3 0.78
tteat 0.76
ass 0.75

This factor accounts for 67 % of the variance in this matrix.

This is the only analysis so far where the QMC #3 has a satisfactory loading. As this analysis contains only verbal tests, this means the QMC #3 catches in only verbal ability and lacks loading on the other ability types.

Test	ass	an	fin	tteat	gat	qmc3
ass	1	0.69	0.75	0.45	0.52	0.25
an	0.69	1	0.77	0.64	0.69	0.46
fin	0.75	0.77	1	0.81	0.72	0.6
tteat	0.45	0.64	0.81	1	0.2	0.64
gat	0.52	0.69	0.72	0.2	1	0.87
qmc3	0.25	0.46	0.6	0.64	0.87	1

Analysis #4 - Verbal, numerical, spatial - March 2006

fin 0.91
num 0.84
sth 0.81

This factor accounts for 73 % of the variance in this matrix.

Test	fin	num	sth
fin	0.93	0.71	0.68
num	0.71	0.86	0.57
sth	0.68	0.57	0.83