Combining psychometric data on brain damage and the influence of aging

60 young and middle-aged patients (age 20 to 52 yr.) with generalized and chronic brain damage were individually matched for age, social class, education, and occupational history with normal control subjects with no history of brain damage. Both groups were given a battery of cognitive tests which included Clément's Digit Code test, Memory-for-Designs, Trail Making, the Spiral Aftereffect, Elithorn's mazes, Mill Hill vocabulary, Progressive Matrices, and the Synonym test. The discriminative efficiency of each measure was examined individually and in combination using discriminant function analysis. The most efficient individual discriminator was the Digit Code. When this was combined with the Trail Making Part B, correct over-all classification was 80%. No other addition improved discrimination significantly. Thus a battery of tests taking 10 min. to administer was as effective as the larger battery which took more than 2 hr. The gradient of test scores with age was examined cross-sectionally for each measure. For most tests, gradients were larger for the brain-damaged than for the control group.