Further Data on the Stanford-Binet VIII- and IX-Year Tests
- Author:
Wallace Wallin
Division of Special Education and Mental Hygiene, Department of Public Instruction, State of Delaware The discussions of Madden and Stoke 1 recall similar statistical studies by tlie writer of the responses to the individual tests in the Binet scale. These studies include an analysis in 1911 of successful responses to the tests in the 1908 scale, based on the performance of 333 epileptics;2 a tabulation in 1916 of the percentage of successes for each test in the 1908 scale based on the responses of 1,000 consecutive clinic cases when the responses were classified (a) according to Binet age irrespective of chronological age and brightness classification, and (&) according to brightness classification irrespective of Binet age or chronological age;3 and a detailed analysis from 1926 to 1928 of successes on the tests in the Stanford-Binet scale. The results of the Stanford-Binet analyses have been published only for the VIII- and IX-year tests, based on 1,382 consecutive clinic cases, of whom over one-fourth were classified as normal mentally, almost one-half as subnormal (borderline and backward), and almost one-fourth as mentally deficient.4 The data have been classified in a considerable variety of ways in order to measure the influence of various factors.5
1 E. Madden, A note on the eight and nine year levels of the StanfordBinet, School and Soc., 1932, 36, 576. M. Stoke, The eight and nine year levels of the Stanford-Binet scale, School and Soc., 1933, 37, 459-461. 2 J. E. Wallace Wallin, Experimental studies of mental defectives: A critique of the Binet-Simon tests and a contribution to the psychology of epilepsy, Baltimore, Warwick & York, 1912; The mental health of the school child, Yale University Press, 1914, pp. 182-208.
3 J. E. Wallace Wallin, The individual tests in the Binet-Simon scale, Psychol. Clin., 1917, 11, 79-85. Only the results for the weight test were published. 4 J. E. Wallace Wallin, A statistical study of the individual tests in ages VIII and IX in the Stanford-Binet scale, Ment. Meas. Monog., 1929, No. 6. s However, the material that has appeared in print was greatly abbreviated for the sake of economy of printing by an assistant of the Mental Measurement Monographs. The condensation and alteration eliminate some of the more detailed and subtle analyses of the original. The original tabula94 The following comments will be limited to a comparison of Stoke’s, Madden’s and my own results for the individual tests in the VIII- and IX-year levels of the Stanford-Binet. Stoke’s data are based on 133 seven-year olds selected from homes which are “fairly representative of American life, except for the fact that there are no farmers or skilled labor representatives,” and Madden’s on “92 children in a poor section of Manhattan,” whose average chronological age was ll^i, average Binet age 9?1, and average I.Q. 80.25. My data are based on the block of clinic cases that was diagnosed as normal on the basis of the synoptic findings from a fairly complete clinical examination. For the Binet VHI-year level only children are included who were eight years old chronologically and those who had a Binet test age of VIII (about 75 cases), and for the Binet IX-year level only those are included who were nine-years old chronologically and those with a Binet age of IX (about 60 cases). A few comparisons will, however, be
Table I Percentage op Successes in the Binet VIII-year Tests Stoke Madden * Wallin, Normal 8-year olds Wallin, Binet VIII-year olds Ball and field Counting backwards Comprehension questions.. Similarities Definitions Vocabulary 14 53 47 39 53 11 41.3 94.6 66.3 56.5 79.3 42.4 54.4 69.4 74.6 68.7 83.3 66.2 58.1 74.7 89.3 71.2 88.6 59.2
Madden’s figures have been converted into percentages on the assumption that all of the 92 children were given each test.
Table II Percentage of Successes in the Binet IX-year Tests Stoke Madden * Wallin, Normal 9-year olds Wallin, Binet IX-year olds Giving date Discriminating weights. Making change Reverse digits Sentence construction. . Rhymes 19 15 36 23 29 49 62.0 44.6 88.0 69.6 71.7 51.1 65.6 60.3 74.6 67.2 86.7 77.9 67.2 66.1 74.5 68.5 92.5 74.5
tion for the different Binet ages can be made available to anyone who desires to analyze the data. made between the results for these normal groups and the results for subnormals and mental deficients.6 A study of the percentages of successful responses in tables I and II reveals the following.
1. The results agree in showing a large range of difference between the highest and lowest percentage of passes in each age level, as shown by the following figures: for the VHI-year tests, Stoke, 42 per cent; Madden, 53.3 per cent; and Wallin, chronological series, 28.9 per cent; and Binet series, 31.2 per cent; for the IXyear tests, Stoke, 30 per cent; Madden, 43.4 per cent; and “VVallin, both series 26.4 per cent. The range is the greatest in Madden’s and the least in Wallin’s series.
2. In spite of certain discrepancies, the results of the three investigations are in fairly close agreement on certain tests. All find the ball-and-field and vocabulary tests the most difficult in the VHI-year level, and the weight test in age IX. Except for a onestep difference in the case of Madden, the date and reverse digit tests rank second and third in difficulty. All find the definitions and sentence construction tests relatively easy. The order of difficulty is almost the same for all the tests in both of Wallin’s series. 3. All three investigations agree in finding the ball-and-field and vocabulary tests harder than any IX-year test, although this applies to only the Wallin Binet age series so far as the vocabulary test is concerned. Making change is easier than five tests in age VIII according to Madden’s and four according to Wallin’s results. The sentence test is easier than any VHI-year tests in Wallin’s results and easier than four YHI-year tests in Madden’s results. VHI-year Tests
IX-year Tests Counting Backward; Stokes Rhymes; Stokes, Madden Comprehension Questions; Stokes Weights; Madden (?) Definitions; Stokes Similarities; Madden (?) Ball and Field; Wallin (Chronological age)
6 For the detailed performance of children of other ages (from chronological and Binet ages four to sixteen) and of subnormals and mental deficients, see the original monograph. 4. On the assumption that a test passed by 50 per cent of the pupils of a given chronological age is standard for that age (the assumption on which the Stanford-Binet tests are standardized) the preceding tests are placed approximately correctly according to the findings of the different investigators. The following tests are too difficult: VHI-year Tests IX-year Tests Ball and Field; Stokes, Madden Weights; Madden (?) Similarities; Stokes All except Rhymes; Stokes Vocabulary; Stokes, Madden (?) The following tests are too easy: VIIIvear Tests IXyear Tests Counting Backward; Madden, Wallin (Both series) All; Wallin Comprehension Questions; Madden, Wallin (Both series) Definitions; Madden, Wallin (Both series) Similarities; Wallin (Both series) All except Weights and Rhymes; Madden Vocabulary; Wallin (Both series) None (?); Stokes None; Stokes 5. On the 66.3 standard of passing, which may seem preferable from certain points of view, the following tests are approximately correct: Vlll-year Tests IX-year Tests Comprehensions Questions; Madden Date: Madden, Wallin (Both series) Counting Backward; Wallin (Chronological age) Reverse Digits; Madden, Wallin (Both series) Similarities; Wallin (Both series) Sentence Construction; Madden (?) Vocabulary; Wallin (Both series) Weights; Wallin (Chronological age) None; Stokes None; Stokes 98 THE PSYCHOLOGICAL CLINIC The following are too hard: VHI-year Tests IX-year Tests Ball and Field; Stokes, Madden; Wallin (Both series) Weights; Madden, Stokes, Wallin (Chronological age?) Similarities; Madden (?), Stokes Rhymes; Madden, Stokes Vocabulary; Madden, Stokes, Wallin (Binet age) All; Stokes All; Stokes The following are too easy: VHI-year Tests IX-year Series Counting Backward; Madden, Wallin (Binet age) Making Change; Madden, Wallin (Both series) Comprehension Questions; Wallin (Both series) Sentence Construction; Madden (?), Wallin (Both series) Definitions; Madden, Wallin (Both series) Rhymes; Wallin (Both series) None; Stokes None; Stokes
6. It is obvious from the data presented above that the tests for these ages at least (VIII and IX) in the Stanford-Binet are in need of thorough revision. As early as 1918 after daily use of the scale in the clinical examination of many hundreds of cases, the writer issued a plea for the thorough revision of the scale not only so far as concerns the accuracy of the standardization but also the administrative procedure of many tests.7
7. In spite of the apparent inaccuracy of the VHI-and IX-year tests, the analysis we have elsewhere made seems to show that most of these tests possess considerable differential value.8 8. One of the most interesting discoveries in our analyses of the test results for the VIII- and IX-year tests may help to explain the discrepancies in the experimental findings of Madden, Stoke, and Wallin and similar discrepancies which frequently occur in intelligence tests findings, namely the fact that subjects 7 J. E. Wallace Wallin, Preliminary impressions of the Stanford revision of the Binet-Simon scale, Psychol. Clin., 1918, 12, 1-15. 8 J. E. Wallace Wallin, A statistical study of the individual tests in ages VIII and IX in the Stanford-Binet scale, p. 13. of the same intelligence (Binet) level but of different degrees of brightness, and younger and brighter subjects and older and duller subjects of the same Binet age may react quite differently to the same test. Thus our younger and brighter subjects did consistently better in the ball and field, the comprehension questions, and the weights, and, in the main, in the rhymes and definitions tests. Older and duller children surpassed the younger and brighter ones in making change, counting backward, size of vocabulary, and, although less decidedly, in the date and the reverse digits tests. Life, age and experience exert a positive influence on some tests, in consequence of which younger and older children of the same Binet level react differently. The explanations obviously differ according to the nature of the test. Thus the older children have more experience in making change and have acquired a richer vocabulary through wider exposure. Perhaps sensory discrimination in kinesthetic sensibility is more acute in the younger children, while younger children doubtless are subjected to rhyming exercises in school more frequently than are the older ones. The reactions to psychological tests may differ in a similar way for normal, subnormal, and mentally deficient children of the same intelligence (Binet) level. Thus, although the ball-and-field test is the hardest for all groups, it is disproportionately difficult for the mentally deficient and subnormal, hence it may possess considerable diagnostic value. Definitions is next to the easiest test for the normal and subnormal, but next to the hardest for the feebleminded. The vocabulary test is easiest for the normal, and hardest for the mentally deficient. Counting backward is easiest for the normal children, and easier for the mentally deficient than for the subnormal. Rhyming is relatively easy for the normal and difficult for the subnormal. Limitation of space does not permit us to point out somewhat similar differences in the reactions of epileptics.9 9. Enough facts have been presented to show that the composition of the group is an important factor to consider in the effort to establish age norms for psychological tests. ‘’ Unselected” groups of children on which norms are usually based, which are assumed to represent symmetrical distributions of cases, may, in 9 See, however, J. E. Wallace Wallin, Experimental studies of mental defectives, a critique of the Binet-Simon tests and a contribution to the psychology of epilepsy; The measurement of mental traits in normal and epileptic school children, Miami Univ. Bull., 1923, Ser. 21, No. 8; and Problems of subnormality, Yonkers, N. Y., World Book Co., 1917, pp. 350-381. point of fact, be based on highly skewed distributions, more especially when the standards are based on a very limited selection of subjects.10 From our data it appears that the norms will be unequally affected for different tests according as the groups are weighted with inferior or superior subjects. The “unselected” samplings on which many norms for psychological tests have been established, have, no doubt, given results that have been skewed too high or too low, because the cases have not represented a symmetrical distribution of mental ability. Moreover, that reliable standards can be established without wide-range testing is seriously to be questioned because of the role of the element of individual variation in psychological test findings. The marked individual differences which occur with the use of the Binet scale is indicated by the following: some of my Binet IV-year olds pass the comprehension-questions and counting-backward tests. Some Y-year olds pass the reverse-digits, sentenceconstruction, and rhymes tests. Some normal Xll-year olds fail on making change, reverse digits, or discriminating weights. Some normal XHI-year olds fail on the date, weights, or rhymes tests. On the basis of our findings it is reasonable to assume that the Stanford-Binet scale is a less valid measure of intelligence for adults, whether normal or abnormal, than for children whether normal or abnormal.
Must we not give more consideration to the nature of the human samples on whose reactions we attempt to establish national norms ?11 10 The Standard-Binet VHI-year tests were standardized on 100 eightyear olds, and the IX-year tests on 113 nine-year olds. (This statement will serve to correct an erroneous statement made on page 53 of the Mental Measurement Monograph referred to on an earlier page.) 11 It is of interest to note that we found no very significant sex difference in our study of these twelve tests.
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.