The Constancy of Intelligence Quotients with Borderline and Problem Cases

• Published in: Journal of educational psychology 1923-04, Vol.14 (4), p.247-250
• Main Authors: Henmon, V. A. C, Burns, H. M
• Format: Article
• Language: English
• Subjects: Borderline States; Human; Intelligence; Intelligence Quotient
• ISSN: 0022-0663; 1939-2176
• DOI: 10.1037/h0070797

Seventy-two pupils referred to the mental examiners as a consequence of their difficulties encountered were twice tested, at intervals ranging from nine months to four years, and the results tabulated. The tests used were the Goddard and Stanford revision of the Binet scale. The median I. Q. for the group was 74, the range from 60 to 90. Where the Binet test was used first and for the retest the correlation found was .91, with an average difference of 5.3 points I. Q. For eighteen cases where the Goddard test was the first given and the Stanford was used for the retest, the average difference was nine points I. Q. The probable error of measurement for the whole study was found to be about three points I. Q. The typical change found in this study was a loss for the retest. It is well established, say the authors, that with feebleminded subjects the I. Q.'s tend to decrease. This is the evident tendency with borderline cases, too, and should be held in mind in making provisions for them or predictions concerning them. No sex differences were to be found in the results of this study, and no differences were shown relating to the different intervals between test and retest. From Psych Bulletin 20:09:00995.