The Correlation of the Reverse Audito-Vocal Digit Memory Span With the General Intelligence and Other Mental Abilities of 308 Prisoners of the Eastern State Penitentiary of Pennsylvania
- Author:
Franklyn D. Fry, University of Pennsylvania
The purpose of this study is to determine the amount of correlation, if any, between the reverse audito-vocal digit memory span and the following: (1) General intelligence, (2) Vocabulary, (3) Interpretation of fables, (4) Chronological age.
The Intelligence Quotient obtained by means of the Stanford Revision of the Binet-Simon Tests was used as a measure of general intelligence. The reverse audito-vocal digit memory span, the vocabulary, and the ability to interpret fables were measured by the individual tests for these abilities contained in the Binet-Simon series. The procedure in administering these tests and the method of scoring them were the standardized procedure and method described in “The Measurement of Intelligence.”1 The subjects of this investigation were originally 507 prisoners, 331 white and 176 colored, of the Eastern State Penitentiary of Pennsylvania. Thirty-five of these prisoners were tested by myself while the remainder were tested by Mr. J. D. Shearer, Psychologist and Director of Education of the Penitentiary. Of the total number of cases investigated, however, 199 were rejected because they had not passed a reverse audito-vocal digit memory span test. For example, if a subject’s basal age should be ten and if he should fail to pass the reverse audito-vocal digit memory span test for five digits at the twelve-year level, then the record would show what he could not do and not what he could do with respect to the reverse audito-vocal digit memory span since there is no test for this ability at the ten-year level. It might be assumed that he could have passed the test for four digits in reverse order which is located at the nine-year level, but this would be an assumption and not an established fact. The study is thus confined to 308 cases of whom 227 were white and 81 colored.
1 Terman, L. M., The Measurement of Intelligence. Now York: Houghton Mifflin Co., 1916.
The score for the reverse audito-vocal digit memory span was the number of digits which could be produced correctly in reverse order on one presentation of the stimulus. The vocabulary score was the number of words correctly defined from the given list of one hundred words. The score for the interpretation of fables was two points for each fable correctly interpreted and stated in general terms, and one point for each fable correctly interpreted but not stated in general terms. Since all of the subjects were adults the nearest birthday was taken as the chronological age. The results for the various tests and for the chronological age are shown in the following tables.
Table i Keverse Audito-Vocal Digit Memory Stan White Colored Total Maximum 7 7 7 75% 6 5 5 Mode 5 4 5 Median 5 4 5 Mean 4.96 4.65 4.88 25% 4 4 4 Minimum 3 4 3 S. D 1.036 0.804 0.989 V 20.904 17.275 20.281 Table ii Intelligence Quotients White Colored Total Maximum 118 107 118 75% 91 79 88 Mode 71 & 75 59 59 Median 79 67 76 Mean 80.9 69.9 77.9 25% 71 59 66 Minimum 48 48 48 S. D 15.822 13.217 15.915 V 19.576 18.891 20.412 158 THE PSYCHOLOGICAL CLINIC Table iii Vocabulary White Colored Total Maximum 90 70 90 75% 64 48 58 Mode 48 30 48 Median 48 37 46 Mean 51.7 39.3 48.4 25% 41 30 36 Minimum 17 16 16 S. D 16.558 12.510 16.519 V 32.048 31.835 34.192 Table iv Interpretation of Fables White Colored Total Maximum 8 8 8 75% 8 4 7 Mode 0 & 8 0 0 Median 4 1 4 Mean 3.9 2.4 3.5 25% 0 0 0 Minimum 0 0 0 S. D 3.209 2.639 3.135 V 82.493 107.569 89.316 Table v Chronological Age White Colored Total Maximum 70 73 73 75% 38 34 37 Mode 25 22 22 Median 28 30 29 Mean 31.7 30.7 31.5 25% 24 24 24 Minimum 16 18 16 S. D 10.624 9.130 10.247 33.462 29.759 32.564
Table VI shows the percentage of eases which were superior, normal and subnormal for each test. It is to be noted that the terms “normal” and “subnormal” are used in a very specific and not in a general sociological sense. Normal cases pass the tests at the “Average Adult” level; subnormal do not.
Table vi Percentages Ability Score Class White Colored Total R. M. S. 7 Superior 10.6% 6.2% 9.4% 6 Normal 15.4 3.7 12.3 0-5 Subnormal 74.0 90.1 78.3 Q. 110-119 Superior 2.6 0.0 2.0 90-109 Normal 25.8 11.1 20.5 0-89 Superior 11.5 0.0 8.4 Voc. 75-100 Superior 11.5 0.0 8.4 65-74 Normal 11.5 2.5 9.1 0-64 Subnormal 77.1 97.5 82.5 I. F. 8 Superior* 29.1 11.1 24.4 & Normal 0-7 Subnormal 70.9 88.9 75.7
There is no test for subjects of superior intelligence in the interpretation of fables therefore this score includes both the superior and the normal.
The proportion of these subjects who are subnormal in regard to the different abilities tested is a clear demonstration of the relation between mental retardation and delinquency. In Table VII will be found the percentages of cases falling in the various classes of intelligence according to Terman’s classification. It will be noted that 22% of the white and 59% of the colored are definitely feebleminded if judged solely on the basis of the Intelligence Quotient.
Table vii Percentages of General Intelligence I. Q. Intelligence White Colored Total 110-119 Superior 3.0% 0.00% 2.3% 90-109 Normal 23.4 11.1 20.1 80-89 Dullard 21.2 13.6 19.2 70-79 Borderline 30.0 16.0 26.3 Below 70 Feebleminded 22.4 59.3 32.1
The colored group of prisoners have a lower score in every test than have the white group. This is due not to racial inferiority as many believe but to the fact that racial prejudice prevents the colored race from entering upon tasks which are too difficult for the higher grade of colored defectives. The higher grade of colored defectives can carry on the tasks in which they are permitted to engage by society and satisfy their simple desires with their earnings without resorting to illegal means as do the whites in their at160 THE PSYCHOLOGICAL CLINIC tempt to satisfy their desires and to maintain a higher standard of living. Industry has not as yet availed itself to a very large extent of the advantages offered by psychological tests in selecting employees for positions according to their special and general mental fitness. Consequently many of the whites secure positions which require a higher level of mentality than they possess, and when desires together with the opportunity afforded by these positions for the gratification of such desires are encountered, these persons sooner or later fail to comply with the regulations of society and fall into the toils of the law. This explains in part why 75.3% of the colored and only 52.4% of the white are of borderline or feebleminded intelligence.
The colored below the borderline level of intelligence must compete in the labor market not only with the white of an equal level but also with the colored of a higher level of intelligence who are forced into such work because racial prejudice prevents them from entering the fields of endeavor in which they are competent to work. Therefore the colored laborers of the lower levels of intelligence are the first to be forced into unemployment and then when they are unable to get employment they attempt to secure a livelihood through illegal means. According to the “Philadelphia Tribune” 2 68% of the total number of the colored employed in Philadelphia during 1925 were engaged in unskilled labor. Mr. S. L. Wade,3 Secretary of the Research Department of the Armstrong Association of Philadelphia, found in the survey which he completed in May, 1929, that 89.4% of the total number of the colored employed in Philadelphia were engaged in unskilled occupations. In Table VIII are given the coefficients of correlation calculated by the Pearson product moment method.
For the colored group there is a high degree of correlation between general intelligence and the reverse digit memory span, vocabulary and interpretation of fables, but for the white group there is a high degree of correlation between general intelligence and the reverse digit memory span only. Vocabulary depends both upon native ability and education. The fables used in the interpretation of fables test are read to the subject therefore there are three factors which operate in the interpretation of the fables. These 1 The Philadelphia Tribune Publishing Co., Philadelphia. ? Mr. S. L. Wade, Secretary of Research Department of Armstrong Association, Philadelphia.
Table VIII. Coefficients of Correlation Correlatives White P.E. Colored P.E. Total P.E. R.M.S. R.M.S. R.M.S. R.M.S. I.Q. I.Q. I.Q. Voc. Voc. I.P. I.Q. Voc. I.F. C.A. Voc. I.F. C.A. I.F. C.A. C.A. 0.736 0.478 0.475 -0.144 0.839 0.744 ?0.065 0.674 0.072 -0.088 ?0.021 ?0.035 ?0.035 ?0.064 ?0.013 ?0.020 ?0.045 ?0.024 ?0.045 ?0.044 0.829 0.744 0.672 0.178 0.792 0.819 0.169 0.756 0.333 0.207 ?0.023 ?0.033 ?0.041 ?0.073 ?0.028 ?0.025 ?0.073 ?0.032 ?0.067 ?0.072 0.752 0.535 0.524 -0.073 0.846 0.770 -0.003 0.704 0.129 -0.019 ?0.016 ?0.027 ?0.028 ?0.038 ?0.011 ?0.016 ?0.038 ?0.019 ?0.038 + 0.038
three factors are vocabulary, the ability to deal with abstractions and the ability to make generalizations the latter two being largely native abilities. Thus the results in both the vocabulary and the interpretation of fables tests depend upon native ability and the educational factor. For this reason the coefficients of correlation of both vocabulary and interpretation of fables with the reverse digit memory span are strikingly parallel. The cause for the coefficient of correlation between general intelligence and vocabulary being greater for the white group and less for the colored group of prisoners than is the coefficient of correlation between general intelligence and either the reverse digit memory span or the interpretation of fables is due to the influence of education. The white group have had the advantages of good educational facilities and a compulsory attendance law while the colored on the other hand have had very few advantages of education since the majority of the colored have migrated from the South where little or no educational facilities are provided for the colored race. This condition exists in the city of Philadelphia, especially since much of the increase in the colored population of Philadelphia for the past fifteen years has been due to migration from the South. The coefficient of correlation of the reverse digit memory span with general intelligence is rather high for both the white and the colored, +.736 ?.021 for the white +.829 ?.023 for the colored. Thus the coefficient of correlation of the reverse digit memory span, a native ability, with general intelligence is .093 higher for the colored than is that of the white. The coefficient of variability of the reverse digit memory span is .328 greater for the white but .616 less for the colored than is the coefficient of variability of general intelligence which is due to the fact that there are superior adults in the white group while there are no superior adults in the colored group. Taking both groups as a whole the coefficient of variability of the reverse digit memory span is .131 smaller than that of general intelligence, while the correlation for the reverse digit memory span with general intelligence is rather high, +.752 ?.016.
With respect to the reverse digit memory span and vocabulary there is a high degree of correlation for the colored group, + .744 ?.033; a decided degree of correlation for both groups, taken as a whole, which is +.535 ?.027; and a low degree of correlation for the white group, which is +.478 ?.035. This difference of .266 between the coefficients of correlation of the reverse digit memory span with the extent of vocabulary for the white and the colored groups is due to the educational factor discussed above. When the reverse digit memory span is compared with the interpretation of fables the coefficient of correlation closely parallels, being slightly lower than that of the reverse digit memory span with vocabulary. This is true for the white, the colored, and for the group as a whole.
There is no correlation between chronological age and vocabulary because these subjects have not availed themselves of experiences essential to the increase of vocabulary. There is no correlation between chronological age and either the reverse digit memory span or general intelligence because these are native abilities which have reached full maturity in the adult. The interpretation of fables depends both upon vocabulary and native ability which latter has likewise matured by the time adult life is attained.
Some contend that the memory span is affected by an individual’s imaginal type. Ilumpstone says,4 “Some imagination is required. The subject must have enough imageability to get perceptions of the stimuli. If auditory imageability, for instance, were ‘Ilumpstone, H. J. Some Aspects of the Memory Span, Philadelphia: The Psychological Clinic, Press 1917. entirely wanting, there could be no response to the auditory presentation.” MeCaulley states, “Likewise, in giving digits in reverse order several methods are used. Some individuals depend on their good visual imagery, and by closing their eyes, or staring raptly before him, they rapidly ‘read’ the digits in reverse order with great facility Other individuals depend largely upon their kinaesthetic-verbal imagery, saying the number forward, then backward, and, if a series is complex, breaking it into groups, reversing it into groups, reversing one group at a time. Still others will ‘write’ the digits with their fingers, or touch each finger to indicate a digit.”5 I have found similar cases in my investigation of the prisoners at the Eastern State Penitentiary of Pennsylvania. According to Thorndike and Wallin the majority of individuals do not, however, show so emphatic a predominance of one kind of imagery as to be put surely in one class. They are mixed types and the extreme audiles, visuals, or motiles are very rare. Therefore the memory span is likely not to be affected to any appreciable extent by an individual’s imaginal type. The use of the reverse digit memory span as a reliable intelligence test does not seem to be justified by experiment. Brotemarkle6 states that, “The diagnostic value of the memory span test is readily evidenced in the manner in which it clearly differentiates the two extreme groups, the upper and the lower. The memory span test is diagnostic of the complexity of mental organization taken as a bodily factor, including the competencies with which the individual makes for successful performance The memory span test has a certain restriction placed upon its use by the intellectual level of the individual being tested. This is clearly noted in the varying diagnostic values of the different types of memory span. Further research in this field should reveal the relative diagnostic use of the different types of memory span tests at the varying intellectual and age levels.” The reverse digit memory span has diagnostic value in differentiating the two extreme groups, the upper and the lower, but it has no value as an intelligence test which will differentiate by small degrees the intelligence of one individual from that of another.
8 MeCaulley, Selinda. A Study of the Relative Values of the AuditoVocal Digit Memory Span and the Reserve Span as Diognostic Tests. Psychological Clinic, 1928, vol. 1G, pp. 277-291. a Brotemarkle, R. A. Some Memory Span Problems. Psychological Clinic 1923, vol. 15, pp. 229-258.
Conclusions
1. There is a high degree of positive correlation between the reverse audito-vocal digit memory span and general intelligence. 2. The coefficient of correlation between the reverse auditovocal digit memory span and general intelligence for the colored group of prisoners is .093 higher than that for the white group. This difference is due to the educational factor which affects some of the tests in the Stanford Revision of the Binet-Simon Scale. 3. The extent of the vocabulary depends both upon native ability and upon the educational factor. When the extent of the vocabulary depends mainly upon native ability there is a high degree of correlation between it and the reverse audito-vocal digit memory span as is the case with the colored subjects included in this study, but when the extent of the vocabulary is highly influenced by the educational factor as is the case with the white subjects included in this study there is a lower degree of correlation between it and the reverse audito-vocal digit memory span.
4. There is a decided degree of correlation between the reverse audito-vocal digit memory span and the interpretation of fables for the colored but a lower degree for the white group of prisoners. The difficulty with the interpretation of fables test is getting the idea of the fable over into the mind of the subject by means of vocabulary which is subject to the influence of the educational factor. This evident fact is demonstrated by a high degree of correlation between the interpretation of fables and vocabulary for both the white and the colored groups; a low degree for the white and a high degree for the colored of correlation between the reverse digit memory span and either the vocabulary or the interpretation of fables; and, the fact that the correlation between the reverse digit memory span and vocabulary for both the white and the colored parallels that between the reverse digit memory span and the interpretation of fables for both the white and the colored.
5. There is no correlation of general intelligence nor any of the more specific mental abilities considered in this study with chronological age because all these abilities have attained their full maturity when adult life is reached.
6. When one result, which is due to a native ability only, is compared with another result, which is due to the combination of a native ability and an educational factor, the degree of correlation varies inversely with the educational factor.
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.