It is usually assumed that although Memory Span is on the whole a poor measure of intelligence and its scores show rather little individual differences, it is nevertheless an extremely good test at the lower levels of mental ability. The present paper reports how Span for digits in WISC (Japanese edition) correlates with intelligence and academical efficiency.
Subjects are: 83 mental defectives (MDs), from imbecilities to borderlines, 6-18 years of age: 263 normals, 6-15 years; 23 cerebral palsied children (CPs), 7-14 years. In this paper academical efficiency is in some cases used absolutely as the proof of intelligence. As an example, we can very safely assume that a child has normal intelligence when he is able to master operations of fractions, decimals, i.e., rational numbers.
Findings and interpretations are: (a) Ability to repeat 6 digits forward and 4 or 5 backward means that the subject has normal intelligence, with very rare exceptions. But not always vice versa. (b) MDs, including borderline adolescents, hardly go beyond 5 digits forward, and 4 backward, with only one exception in this report. Again not always vice versa. (c) Generally speaking, correlations of subtests with global intelligence are lowest at the younger age (4 years in this report), and highest at the adult age, and again a little lower at the senility. While the correlation of Span with global intelligence remains constant throughout the different age levels, scoring about .5-.6. Hence it occurs that at the Kindergarten age, Span is one of the most effective subtests for measuring intelligence, (d) Development of Span of MDs as related to MA shows three stages. Below 4 years of MA, Span does not exceed 3 forward, no backward. Within 5-7 years of MA, Span ranges from 3 forward, and no backward, to 5 forward, and 4 backward. Here Span and MA have no correlation. Beyond 8 years of MA, Span is 5 or 4 forward, and 4 or 2 backward. (e) It is rather rare that a child has superior academical efficiency in spite of a little low IQ. Such a case can be found in CPs who have received enough academical training in spite of many hadicaps in everyday life. And it was found that a superior Span (here 5 or more forward, and 4 or more backward, ) goes with normal academical achievements, that is, the virtually normal intelligence, again with very few exceptions. (f) Span seems to have a particular role in thinking or reasoning. Span is a sort of capacity to retain in mind several elements which have not logical relationships between each other. While in solving a problem, we must often retain in mind several elements before we can make out the logic of them to hit at the correct solution. Moreover, Span means a sort of mental effort to keep attention and to inhibit automatism which often distracts thoughts from the correct course. (g) We can attribute a special validity to such tests as devised by Rey (1955) which requires the capacity to retain in mind a set of nonsense materials or nonsense series of stimuli. (h) Whether Span is subject to training, and if so, whether the increase of Span might be accompanied with that of intelligence or not, remains unsettled.