Evaluation of Kojima-Matsubara color vision test plates: validity in young children

We examined a pseudoisochromatic color plate test by Kojima and Matsubara for young children which uses drawings of familiar objects rather than letters or numbers. First, we evaluated the test's efficacy as a color deficiency screener and its validity in classifying the types of color deficiencies by comparing its results with those from the Moreland anomaloscope. Second, we eliminated the chromatic factor and evaluated the functional ability of young children to perform the task by determining how many correct responses were obtained using modified black/white replicas of the test plates.

Part 1: Twenty color-normal and 13 color-deficient adults were diagnosed and classified with the Ishihara test, Panel D-15 test, and anomaloscope. Subjects were then tested with the Kojima-Matsubara test and result were compared with those from the anomaloscope. Part 2: Fifty children aged 3 to 7 years were tested with modified black/white test plate replicas. The number of correct responses for each plate was determined for five different age groups.

Part 1: Among the 20 color-normal subjects, 18 read all 10 plates correctly and 2 subjects missed 1 of the 10. Only 1 of the 13 color-deficient subjects exhibited the expected responses for plates 2 to 6 (used for color deficiency screening). The color-deficient subjects' responses for plates 7 to 10, which are used to classify red-green defects, were varied and only the protanomalous subjects (n = 2) followed the expected response pattern. Part 2: Of the 10 black/white modified plates, only 2 were correctly identified by all 50 children. The other plates had a recognition rate that ranged from 32 to 98%.

Because the response patterns given by most of the color-deficient adult subjects were different from those in the test manual, ambiguous results would occur if the Kojima-Matsubara test were used for color vision screening or the diagnosis of color deficiency. In addition, the difficulty that many of the young children exhibited in identifying the objects in the black/white replica plates suggests that there would be a large number of false positive errors (classifying a color normal as color deficient) when using this test in young children.