clusive, serious consideration should be given to the criteria for selection of children to these classes, with those of 39 or 40 IQ and above being probably the better prospects. They also recommended that emphasis on academic instruction should be discouraged.

Kolstoe 10 studied the language learning of a group of residential school mongoloid children with an average IQ of 25 and an average MA of 2-2, in comparison with a control group having an average IQ of 23 and an average MA of 2-2. This group, therefore, averaged a level somewhat below the group usually found in classes for the middle range. He found, after intensive individual instruction, some gains (not statistically significant) in the experimental group. He pointed out, however, that the gains were concentrated in the higher group. It would appear, therefore, that the timing of learning tasks, especially in terms of mental age, is an important aspect of teaching these children.

Johnson and Capobianco 11 found that the mean social quotients of both their public day school and residential school children increased somewhat from the fall of 1954 to the spring of 1956. A majority of the curves showed a fairly high initial gain, followed by a plateau or slight decrease after the first school year. They reported a slightly larger increase in social quotient in the higher IQ group. The study by Johnson and Capobianco emphasized the measurement quantitative gains by an experimental group as compared with a control group. On a oneyear language training program they reported no difference between the amount of improvement by an experimental and a control group.12 On a measure of academic knowledge, the largest gains were reported on the highest IQ group, and were progressively less for lower IQ classifications. On the Fels Scale both a public school and an institutional group showed changes in a positive direction over the duration of the study. On the Behavior Check List fairly constant gains were reported except in the group under 25 IQ. They concluded that the improvements that take place under a training program will parallel the growth pattern as indicated by the IQ. They interpreted this to mean that the quotient changes they found in some areas of growth would imply little effect on the longrange programs for the individual in personal, social, and economic adjustment. They concluded also that children with IQ's below 30 derived little gain from a program such as was used in their study. They reported, therefore, that, in their judgment, the problem of the severely retarded remains one of training for self-care and socialization to the maximum extent possible, and a total solution to their problem must involve life planning. It was their belief, as a result of the study, that the learning characteristics of the severely retarded are in marked contrast to those of the less severely retarded.

Cantor and Hottel 13 in a study of discrimination learning also found a higher IQ group (above 50 IQ) showing a performance superior to that of a lower IQ group (below 50).

Johnson, Capobianco and Miller 14 reported that severely retarded children scored more than a year below their mental ages in language achievement on a test of language ability, in both public day school and residential school groups. A year of an intensive language development program failed to show any greater gain in experimental groups than in individually matched control groups.

10 Kolstoe, Oliver P. Language Training of Low-Grade Mongoloid Children, American Journal of Mental Deficiency, 63, 17-30, July 1958.

11 Johnson, G. Orville and Capobianco, Rudolph J.

Research Project on Severely Retarded Children.

Albany, N.Y.: New York State Interdepartmental Resources Board, 1957. 230 p. (mimeo.)

12 See also Johnson, Capobianco, and Miller (14) for a more extensive discussion of this problem.

13 Cantor, Gordon N. and Hottel, J. V. Discrimination Learning in Mental Defectives as a Function of Magnitude of Food Reward and Intelligence Level, American Journal of Mental Deficiency, 68, 300–384, October 1955.

14 Johnson, G. Orville, Capobianco, Rudolph, and Miller, Donald Y. Speech and Language Development of a Group of Mentally Deficient Children Enrolled in Training Programs, Exceptional Children, 27, 72-77, October 1960.

On an articulation test of 11 speech sounds, the initial position of a sound appeared to be most frequently correct, the medial position next most frequently correct, and the final position of the sound least often correct. Omission of a sound was the most common error, followed by substitutions and distortions. On the whole, fewest errors were made by children within the group with the highest IQ range. The average mental age of the groups was about 41⁄2 years at the beginning of the study.

A study such as this points out several matters for the teacher to keep in mind. It shows the marked language retardation in these children as a group.

In the Illinois study 15 it was found that IQ's went up slightly over a year of special class teaching. It was felt, however, that factors such as familiarity of the children with teachers and examiners might have had an effect. Teachers' rating on the Behavior Check List indicated noticeable or marked gains on the part of two-thirds of the children. Gains were greatest in the simpler activities, least in the complex activities. It was pointed out, however, that one year was

probably too short a time to lead to definite conclusions.

Goldstein 16 also reported, on the basis of the Illinois study, that the IQ and SQ remained relatively constant over the 2-year period of the study. On the Behavior Check List the teachers reported considerable progress during the first year, and no progress during the second year.

Johnson, Capobianco, and Blake 17 reported significant increases in score on the Vineland Social Maturity Scale, the Fels Child Behavior Scale, and a Behavior Check List for a group of severely retarded children in day and residential schools once over a two-year period. Gain was primarily during the first year, however.

In some instances, therefore, the relative measures showed initial gains, followed by a leveling off, so that the absolute gains are more nearly those that might have been predicted by some such relative measure as the IQ. Here some caution may be necessary in drawing conclusions. Failure to maintain relative gains, such as increasing intelligence quotient, social quotient, or educational quotient, has sometimes been interpreted to mean that special education of the severely retarded is useless on the basis that the children would have done just as well if they had not attended a special class. Few, if any, educational situations, however, can consistently meet this rigorous criterion. Perhaps the demonstration of absolute gains is a more reasonable criterion. For several reasons, the present studies cannot be considered as being conclusive. The samplings of children have been small, and the duration of the studies has been too brief for the findings to be definitive over a long period of time. Especially with a slow-developing group, the studies need to be continued for a sufficient length of time to show quite definite evidence of terminal status as well. A third difficulty is the fact that the studies have all been based on the methods of instruction in use at the time the study was made. Nevertheless, the results of the studies have been sufficiently clear to indicate that some of the more exaggerated hopes and claims that have been expressed are probably too optimistic.

In summary, practically all the studies have shown some absolute gains, the greater gains, on the whole, having been made by the more able children; the least gains, by the least able. Some of the studies also suggest that there may be quite rapid initial gains, followed by a period of less rapid gain. The gains

15 Illinois State Department of Public Instruction. Report on Study Projects for Trainable Mentally Handicapped Children. Springfield, Ill.; the Department, November 1954, 43 p.

16 Goldstein, Herbert. Report Number Two on Study Projects for Trainable Mentally Handicapped Children. Springfield, Ill.: Illinois State Department of Public Instruction, January 1956. 42 p.

17 Johnson G. Orville, Capobianco, Rudolph J., and Blake, Kathryn A. An Evaluation of Behavioral Changes in Trainable Mentally Deficient Children, American Journal of Mental Deficiency, 64, 881-873. March 1960.

during this later period seem to be most nearly in line with expectancy based on the IQ.

Other Studies of Learning.-Other isolated but relevant experimental findings include the following.

Gordon, O'Connor, and Tizard 18 found that individuals in the middle range were capable of sustained work for more than an hour, and that they responded better to knowledge of results than to competition or the general instruction to do their best. Findings such as these have genuine significance for teaching.

Cantor and Stacey 19 found a relatively sharp break at the IQ level from 42 to 59, this group performing in a much inferior way on the Purdue Peg Board, as a criterion of motor dexterity, when compared with individuals of higher IQ.

Zeaman 20 has recently summarized a series of studies on discrimination learning in this group of children (mental ages 2 to 6). He concluded that they are particularly slow in forming new visual habits, even slower than might be predicted from their mental ages. He reported that his findings were consistent with a theory of attention which holds that these children are more specifically retarded in ability to direct and maintain attention than in learning ability as such. He suggested, therefore, that the attention of these children might be aroused through novel-looking stimuli and that a sequence of tasks going from easy to difficult might be helpful in teaching them.

At least one carefully controlled experimental study 21 has shown that children in the "trainable" group do not perceive and recognize geometrical figures in brief exposures as well as children in the "educable" group. This again suggests, perhaps, that the teacher should give the severely retarded child ample time to "see" objects in a learning situation.

Reducing the problem to its most practical terms, Goldstein" evaluated it in terms of the success of the children in adjusting to the group learning situation. He reported the following percents of children excluded from the program after trial:

On the basis of these data, the chances of success in the special class were good for children with IQ's above 35. Of those with IQ's between 25 and 34, a small percentage survived, while of those with an IQ below 25 practically none survived. Goldstein concluded, therefore, that the critical zone is between 25 and 34 IQ. His report on minimum mental ages showed the following: 71 percent of children with mental ages below 2 years were excluded; 21 percent of those with 2-3 mental ages and 13 percent those with 3-4 year levels were dropped; only 3 percent of the children with mental ages over 4 were excluded. This suggests a critical minimum mental age at the 2- to 4-year mental level.

18 Gordon, S. O'Connor, N., and Tizard, J. Some Effects of Incentives on the Performance of Imbeciles on a Repetitive Tasks, American Journal of Mental Deficiency, 60, 371-377, October 1955.

19 Cantor, Gordon N. and Stacey, C. L. Manipulative dexterity in Mental Defectives, American Journal of Mental Deficiency, 56, 401-410, October 1951.

20 Zeaman, David. Discrimination Learning in Retardates, Training School Bulletin, 56, 62-67, August 1959.

21 Griffith, Ann H. The Effects of Retention Interval, Exposure-Time and IQ on Recognition in s Mentally Retarded Group, American Journal of Mental Deficiency, 64, 1000-1003, May 1060.

22 Goldstein, Herbert. Lower Limits of Eligibility for Classes for Trainable Children. Exceptional Children, 22, 226-227, March 1956.

U.S. GOVERNMENT PRINTING OFFICE: 1961-O 505553

Mentally Retarded

HILL, ARTHUR S. The Forward Look: The Severely Retarded Child Goes to School (Bulletin 1952, No. 11). 54 p. 25 cents.

MACKIE, ROMAINE P., and DUNN, LLOYD M. College and University Programs for the Preparation of Teachers of Exceptional Children (Bulletin 1954, No. 13). 90 p. 35 cents.

State Certification Requirements for Teachers of Exceptional Children (Bulletin 1954, No. 1). 60 p. 25 cents.

and CAIN, LEO F. Professional Preparation for Teachers of Exceptional Children: An Overview (Bulletin 1959, No. 6). 139 p. 35 cents.

WILLIAMS, HAROLD M., and DUNN, LLOYD M. Teachers of Children Who Are Mentally Retarded (Bulletin 1957, No. 3). 97 p. 45 cents.

MARTENS, ELISE H. Curriculum Adjustments for the Mentally Retarded (Bulletin 1950, No. 2). 100 p. 45 cents.

Preparation of Mentally Retarded Youth for Gainful Employment. U.S. Department of Health, Education, and Welfare; Office of Education and Office of Vocational Rehabilitation; and American Association on Mental Deficiency, Project on Technical Planning. (Office of Education Bulletin 1959, No. 28; and Office of Vocational Rehabilitation, Rehabilitation Services Series No. 507.) 86 p. 35 cents.

RICE, MABEL C., and HILL, ARTHUR S. Statistics of Special Education for Exceptional Children, 1952–53 (Biennial Survey of Education in the United States, 1952-54, Chapter 5). 78 p. 30 cents. WILLIAMS, HAROLD M., and WALLIN, J. E. WALLACE. Education of the Severely Retarded Child: A Bibliographical Review (Bulletin 1959, No. 12). 24 p. 15 cents.

WILLIAMS, HAROLD M. The Retarded Child Goes to School. Pamphlet No. 123, 24 p. 15 cents.

U.S. Department of Health, Education, and Welfare, Office of Education. The publications are available from the Superintendent of Documents, U.S. Government Printing Office, Washington 25. D.C.