The Variability of I.Q.’S for Psychopaths Retested with in Fifteen Days
Charles A. Ford
State Bureau of Juvenile Research, Columbus, Ohio This matter of I.Q. variability was a problem much more alive a few years ago than at the present time, yet it still seems to be an open question. We, at the Bureau of Juvenile Research, had stopped reporting I.Q.s, feeling that entirely too much importance was being attached to them. We felt that more was to be gained by a study of the individual and his various performances than by a numerical index. To discourage the attitude of the I.Q. being accepted as a final judgment we stopped using it publicly, although within our own laboratory it was still computed and evaluated according to our other findings. Many attempts have been made in our laboratory to determine important diagnostic factors in the behavior of delinquents. Among the other attempts was one which furnished the data for this discussion. Back in 1923, for a period of some six or eight months, it was made a matter of routine to test all cottage or resident cases two or more times, with the Stanford-Binet. The period between tests was not standard, but varied from one to fifteen days. The examiners were all regular members of the Bureau staff and each had had sufficient clinical experience and training to insure the accuracy of the test within its limits. It is, and was, a matter of routine that the Stanford-Binet be given among the first of the tests. However, the tests were never given until rapport had been established and the child had become accustomed to his new environment. That is to say, the child had been a resident of the Bureau for some variable number of days and was known to the examiner before he was tested. Even then, the Binet test was not the first test given, but there had been at least one shorter and simpler test given as a shock absorber. Each child was examined in the standard way, in a private office, free from distraction. The following are incidental data which have grown out of one of the internal experiments at the Bureau, and are compiled to show the variability of the I.Q. of psychopaths (so diagnosed) who were tested two or three times within a fifteen-day period while residents at the Bureau. Its further purpose is to show the relationship of this variability to sex, chronological age, percentage of psychopathy, and mental level as indicated by a Stanford-Binet I.Q. A total of forty-two cases have been found which fit into this specialized group. Fifteen of the cases are females. These range in age from 5 years and 7 months to 17 years, and in the first I.Q. from 49 to 101. Three of them had been tested three times in the period, the other twelve had been tested but twice. Twenty-seven males are included in the study, and these range in age from 6 years and 8 months to 17 years and 10 months. The first I.Q.s range from 61 to 109. Eight of them had been tested three times in the period, the other nineteen had been tested twice. These data are too few to be given detailed statistical treatment but inspection brings out the main facts. The criterion for psychopathy is the diagnosis of the clinician at the time of the child’s stay at the Bureau. In some cases, this diagnosis was based upon the usual clinical symptoms of psychopathy. These cases are clearly indicated in the tables. Usually, however, the diagnosis of psychopathy means 30 or more per cent of psychopathic function as determined by Dr Mateer’s1 “Ten Points of Psychopathy,’’ which was almost invariably used at the Bureau at that time. This percentage is also indicated in the tables.
Table I Chronological Psychopathic I.Q. I.Q. Age Function First Test Second Test 17-10 50 109 120 (2) 15- 4 50 103 119 (3) 12- 6 40 95 102 (7) 11-10 30 95 93 (2) 14- 8 30 94 90 (9) 7- 8 40 87 92 (7) 13-11 30 85 94 (15) 11. G 60 84 86 (7) 16- 3 50 82 86 (4) 9- 3 40 81 88 (3) 9- 7 30 80 87 (3) 15-11 40 76 78 (8) 17- 3 50 75 68 (7) 15-10 50 73 77 (7) 14- 1 50 66 69 (7) 12- 7 40 64 72 (4) 17- 8 50 63 73 (3) 15- 7 60 62 68 (11) 9-11 30 61 59 (2)
1 Mateer, Florence, The Unstable Child (1924), particularly chapter 10. VARIABILITY OF I.Q.S 201 Table I shows the variability of I.Q.s for boys and Table II, for the girls, where only two tests have been given. These show clearly that the variation is not a function of either the length of time between the tests, the mental level as measured by a Stanford I.Q., or the age of the case. From these data, it seems as though tests of psychopaths that are repeated any time within a fifteen-day period are apt to vary in either direction and in any amount.
Table II Chronological Psychopathic I.Q. I.Q. Age Function First Test Second Test 14- 0 60 101 98 (2) 8- 0 30 98 99 (7) 17- 0 30 86 94 (9) 16- 3 40 73 82 (3) 17- 5 50 71 75 (6) 17- 0 t 70* 92 (1) 16- 8 60 70 78 (7) 11- 2 60 69 69 (7) 16- 7 40 66 68 (8) 14- 4 60 64 68 (3) 16- 8 50 57 58 (3) 16-11 50 49 53 (7)
Note: The figures in parentheses in Tables I and II indicate the number of days elapsing between the two tests. t Amount of psychopathy not expressed as a percentage. * This is a Goddard Binet I.Q.
Table III shows both the boys and the girls who have been tested three times within the period, and who are not included among those in Tables I and II. These cases serve to show further how unpredictable the trend of the I.Q. is for these children, regardless of age or I.Q. on first test. Not only is there a marked variation in test performance between the first and second tests but also between the first and third tests and second and third. Although there are but eleven cases which have had three tests, the variation ranges from minus 4 to plus 14, between the first and third tests, and from minus 13 to plus 13, between the second and third tests. Cases whose I.Q. has increased on the second test may fall on the third test or vice versa, but this is not invariable nor is the rate of increase, when there are two increases, a constant. In short, one cannot predict just what will happen to an individual’s I.Q. by repeating the test three times within a short period. 202 THE PSYCHOLOGICAL CLINIC
Table III Chronological Psychopathic I.Q. l.Q. l.Q. Age Sex Function First Test Second Test Third Test 10-10 F f 104 105 (4) 102 (8) 1G-10 M 40 89 81 (4) 94 (9) 14- 3 M 30 89 89 (3) 90 (5) 11- 4 M 30 88 81 (3) 84 (8) 16- 2 M 30 79 74 (4) 80 (9) 15- 8 M t 74 81 (4) 80 (8) 17- 1 M f 72 75 (2) 86 (15) 4-10 M t 72 62 (1) 69 (3) 14- 4 F 50 67 78 (5) 80 (15) 5-7 F 30 66 66 (3) 74 (10) 6-8 M t 62 83 (3) 70 (6) t Amount of psychopathy not expressed as a percentage. Numbers in parentheses indicate the days elapsing between the tests. It will be noticed that in some of the cases there is no variation or only a variation of one point. This would tend to argue strongly for the constancy of the I.Q. but it is interesting to know that even in these cases, the internal variations of the test are very great. For instance, take the third case listed on Table III whose I.Q. on successive tests ran 89, 89, 90. In arriving at these I.Q.s, the compilation of successes is quite different. The summaries of years and months earned is shown in Table IV.
Table IY Months earned Year Test 1 Test 2 Test 3 8 Basal Basal 9 6 Basal 10 10 10 10 12 12 18 18 18 14 8 4 8 16 15 10 10 Thus it will be seen that only on twelfth year tests did the child have a consistent performance. This case is not the exception but the rule, at least, as far as the psychopaths in this study are concerned. They fluctuate in performance from day to day, and though it may not be evident from their I.Q., it is evident from an analysis of the test itself. However, this paper is concerned only with I.Q. variability, so further discussion of this point will not be made.
Table V shows the frequency of the variations of I.Q.s by sex. VARIABILITY OF I.Q.S 203 These data include the results of the second test as found in Table III, or, in other words, include the entire forty-two cases. It will be seen from this that whereas only one girl, or 6.6 per cent of the female distribution, gets a lower score on the second test, eight, or 29 per cent of the male distribution, lower their score. It will also be seen that six of the males who have decreased their I.Q. have made a larger decrease than the one girl who decreased her score. The largest increase was made by a girl. However, this increase may have been, in part, because the first test was a Goddard-Binet and the second a Stanford. With this one exception, the girls vary between minus 3 and plus 11, whereas the boys vary from minus 10 to plus 21. Thirty-eight per cent of the total distribution, regardless of sex, varies more than five points of I.Q. on the second test, the median of the distribution being at plus 5.
Table V Variation Males Females Total minus 10 to minus 6 4 0 4 minus 5 to minus 14 1 5 0 12 3 plus 1 to plus 5 7 7 14 plus 6 to plus 10 S 3 11 plus 11 to plus 15 1 1 2 plus 16 to plus 20 1 0 1 plus 21 to plus 25 1 1 2 From this it will be seen that the tendency of the group is to increase their I.Q. on a second test. Perhaps some of this effect is due to practice. Yet some of them drop and there is no constancy to the increase so that the practice effect, if any, is a decided variable. The theory of psychopathy is that the condition is one of unbalanced and unstable performance. The test variations seem better explained by this theory than by practice effect. Thus, the variable test performance is simply an objective factor, evidencing the general variability of the performance of the psychopath as theoretically described. If this be true, a fourth of fifth test would be as apt to go up or down as the second or third. We do not have the data to show this, but it is certainly plausible in the light of the performance on the third tests, which do not necessarily follow the trend of variability of the second. Obviously, this only applies to tests given very close together, or, in other words, it is their daily performance which is variable and this can only be shown by repeating tests in short periods of time.
The data show that no definite relationship exists between the percentage of psychopathy and the variation. In each of the groups both increases and decreases are found, and if the data were more extensive, it seems as though the variation might be as great in any one group as any other. Certainly from these data it is evident that the variations do not get larger as the percentage increases and equally small variations are to be found in all the groups.
Conclusions
1. Psychopaths (as diagnosed) vary their I.Q. from minus 13 to plus 22 points when re-tested within a fifteen-day period. Over half of them vary less than plus or minus 5 points, however. 2. A decidedly higher proportion increase than decrease. 3. A higher percentage of boys than girls decrease on second tests.
4. The variability is not a function of age, length of time between tests, percentage of psychopathy, nor mental level as evidenced by the I.Q. on the tests themselves. 5. Practice effect does not explain the variability because of the internal variability of the test performance and lack of uniformity of the effect. 6. The day to day variability is better explained by the theory of psychopathy as unbalance than by practice effect. 7. It is unpredictable which way the second or third test will vary or how much.
Disclaimer
The historical material in this project falls into one of three categories for clearances and permissions:
Material currently under copyright, made available with a Creative Commons license chosen by the publisher.
Material that is in the public domain
Material identified by the Welcome Trust as an Orphan Work, made available with a Creative Commons Attribution-NonCommercial 4.0 International License.
While we are in the process of adding metadata to the articles, please check the article at its original source for specific copyrights.