Reduced changeover intelligence test

from Wikipedia, the free encyclopedia

The Reduced Wechsler Intelligence Test (WIP) is an intelligence test . It was developed by Gerhard Dahl from a statistically justified sub -test selection of the Hamburg Wechsler intelligence test for adults , a German adaptation of the internationally popular American Wechsler scales . A first version of the test was published in 1968, first as Reduced W echsler- I ntelligenztest for P sychiatrisch sick quotes, which stands WIP was introduced. In 1972 and 1986 editions of a WIP manual were published while keeping the abbreviation. Since then, the test has been used to estimate the intelligence quotient using the HAWIE total IQ as an independent and standardized short test.

Test setup

The WIP consists of only four of the eleven subtests of the overall Wechsler test:

  1. General knowledge (AW): 25 questions (political personalities, geography, history, literature, foreign words, etc.) on the "scope of knowledge".
  2. Finding commonalities (GF): Identify commonalities for 12 word pairs for the "ability to abstract and conceptualize".
  3. Adding pictures (BE): Name clearly missing details for 15 pictures about the "(optical) differentiation ability with regard to the understanding of the essentials".
  4. Mosaic test (MT): Reproduction of seven depicted patterns with the help of colored blocks for "visual imagination, combinatorics and constructivity".

The test material includes forms (Form 72 or 86) and additional material from the HAWIE: Test booklet supplement pictures (BE) and the box with 16 mosaic stones in the mosaic test (MT)

With these four subtests, the overall IQ can be reliably determined in a short presentation time of around 15 minutes. If only this statement is required, the entire HAWIE does not have to be carried out and the time burden for clinical patients is reduced ( test economy ). The evaluation takes 5 to 8 minutes.

The two forms of WIP differ in the areas of application:

  • WIP 72 is used as a short form for estimating the HAWIE-IQ for subjects aged 10 - 79 years. In clinical psychology, it is used for the standardized measurement of intellectual performance in neuropsychiatric patients
  • WIP 86 is the statistically further developed form. It is used in the field of advisory aptitude diagnostics, pedagogical or work psychology for intelligence diagnostics in normal-healthy subjects.

For both areas of application, standard values ​​have been determined in the manuals that record intelligence as the overall IQ with regard to validity, reliability and objectivity just as reliably as the complete changer test. Due to the guidelines of DIN 33430 , these standards must now be viewed as outdated, so that a control / revision is urgently required.

development

Some studies with the complete changer test have shown that a differentiated interpretation of performance fluctuations in the sub-tests in the sense of a profile analysis or as an index such as the calculation of the so-called degradation quotient for diagnosing a brain-organic degradation of intelligence are problematic or statistically not tenable.

The Wechsler scales are too extensive and uneconomical to only determine an overall IQ value . If carried out for up to 90 minutes, they can represent an unnecessary additional burden for patients. For these reasons, suggestions for a short form have been made early on, but without any more differentiated statistical analyzes

From the frequent criticism of the inadequacies and methodological errors of the HAWIE, since 1968 and in the following years, partly also with the support of the German Research Foundation, the possibility of a test-statistically secured and economic short form was investigated in order to discuss a methodology that would be used in a revision the HAWIE total scales could be applicable.

The procedure is based on the principles of classical test theory. The HAWIE subtests for the procedure were selected using the Wherry-Doolittle method, a step-wise linear regression procedure , in order to enable an estimation of the HAWIE total IQ with fewer subtests and an acceptable loss of validity. For this purpose, the four sub-tests AW, GF, BE and MT were found to be sufficient from the data of the HAWIE overall test on a patient collective (n = 305) from the Berlin University Nervous Clinic. According to the clinical examination group, this battery of four was provisionally referred to as WIP - Reduced Wechsler Intelligence Test for the Psychiatric Ill . " Although the test has meanwhile also been made available for non-clinical, ie normal population groups, the old abbreviation was retained because it became common practice as a result of the rapid spread of the test " (Brickenkamp, ​​1975, p. 169)

The correlations of the WIP subtest values ​​with the HAWIE total IQ were r = 0.888. Check-ups on a large group of patients (n = 617) at the University Clinic Freiburg i.Br. with an r = 0.901 and several normal groups (r = 0.92 to r = 0.96) fully confirmed the results. They underline the usefulness of WIP for diagnosing general intellectual performance. Other authors found " that the loss of information through the use of the short form compared to the HAWIE is very low " (Baxa et al. 1970, pp. 228-231).

It should be noted, however, that high correlations do not mean that the IQ values ​​determined are numerically identical. Linear shifts do not reduce the correlation. When used as a short form for estimating the total test IQ, Orgass and Hartje determined a standard error of 8.45, which corresponds to a mean deviation of approx. ± 9 points for the IQ between HAWIE and WIP72. But the success of the 1972 WIP does not depend on the numerical agreement with the HAWIE-IQ. Because the test is "more than a time-saving short form ... (er) could well prove to be a further development and improvement of the HAWIE. To check this," Orgass and Hartje go on to say, "the WIP should be completely independent Treat the test and examine its diagnostic properties directly rather than judging it by its compliance with the HAWIE ".
Further statistical studies on different populations made it possible to develop the WIP86 as an independent test with its own standard values.

Quality criteria

Differentiated information on the extensive and subtle statistical analyzes can be found in the WIP manuals. Brickenkamp gives a comprehensive overview:

Objectivity - evaluation objectivity

Six diagnosticians calculated objectivity indices for the four subtests between r = 0.956 (BE) and r = 0.993 (MT) from the independent evaluation of 50 identical test protocols.

Reliability

"Item analyzes" showed different indices of difficulty, so that the arrangement for the WIP had to be changed. The coefficients of a "consistency analysis" according to Kuder-Richardson , formula 20, are between r = 0.53 and r = 0.89 in two random samples of patients and normal healthy individuals for AW. For BE the values ​​are between r = 0.72 and r = 0.53. The internal consistency of GE is between r = 0.44 and r = 0.87 and that of MT between r = 0.73 and r = 0.87. The values ​​were determined by analysis of variance according to HOYT UND STUNKARD. For the entire WIP, the determination of the consistency using the odd-even method with correction according to SPEARMAN-BROWN yielded r = 0.94. Intercorrelations of r = 0.35 to r = 0.57 were found for homogeneity.

validity

For the "internal validity", an average correlation of the subtests with the total WIP of r = 0.62 was calculated. The multiple correlation between WIP subtests and the HAWIE total scale was r = 0.89 and r = 0.97 in two samples. The correlation between WIP and ACTUAL was r = 0.83. Double cross-validations using regression equations from WIP and HAWIE subtest values, one-sided and reciprocal values ​​between r = 0.92 and 0.96 with a standard error of s = 5.8 to 7.0 (Seydel, 1972). For the external validity, the WIP was correlated with the following tests: Test d2 r = 0.38; Benton test r = 0.35.

Standard values

The raw values ​​of all subtests in the Wechsler scales are - as initially also in the WIP - not normally distributed. A linear standard transformation into value points leads to distorted and unrealistic IQ values ​​in the high ranges: According to this, Einstein would have an IQ between 160 and 180, the mathematician Gottfried Wilhelm von Leibniz would have an IQ of 205 above that, but would be Goethe's with an IQ Overtaken by 210 as the most intelligent personality of all time thanks to incorrect test standardization. In the WIP, the abnormally distributed raw values ​​were transferred via the cumulative percentage frequency (WIP 72) or area transformed (WIP 86) into normalized T values ​​or T * values ​​in order to avoid such distortions. But this could not be transferred to the German processing of the Wechsler scales in the HAWIE. Because the "American publisher insisted that the German version should be based on the American original as closely as possible ...", writes the editor of HAWIE-R, Uwe Tewes. The restriction also applies to standardization, and so the foreword of the German version goes on to say: The American “Copyright sets limits on efforts to do justice to all public criticism of the original version in the new version” (Tewes, 1990).

Comparable developments

Such a short test “WIPKI” with the same sub-test selection was also developed by Ingrid Baumert in 1973 for the Hamburg-Wechsler intelligence test for children (HAWIK) and based on the concept and the abbreviation of the WIP.

Individual evidence

  1. ↑ D. Wechsler: The measurement of the intelligence of adults. Text volume on the Hamburg Wechsler intelligence test for adults (HAWIE). 3rd, unchanged edition. Huber, Bern / Stuttgart 1964.
  2. ↑ Gerhard Dahl: Compliance validity of the HAWIE and development of a reduced test form. (= Psychologia Universalis. Volume 14). Hain Verlag, 1968.
  3. ↑ a b c Gerhard Dahl: WIP - reduced Wechsler intelligence test - application, evaluation, statistical analysis, standard values. Publisher Anton Hain, Meisenheim am Glan 1972.
  4. ↑ a b c d Gerhard Dahl: WIP. Manual for the reduced changer intelligence test. Application, evaluation, statistical analysis, standard values. 2nd, completely revised and enlarged edition. Publisher Anton Hain, Meisenheim am Glan 1986.
  5. ↑ a b c R. Brickenkamp: Handbook of psychological and educational tests. Hogrefe Verlag, Göttingen / Toronto / Zurich 1975. (books.google.ch)
  6. ↑ E. Brähler, H. Holling, D. Leutner, F. Petermann et al. (Eds.): Brickenkamp manual of psychological and educational tests. (Vols. 1-3). Hogrefe Verlag, Göttingen / Toronto / Zurich 2002, ISBN 3-8017-1440-3 .
  7. ↑ WIP - reduced changer intelligence test (PSYNDEX Tests Review) on zpid.de
  8. ↑ G. Dahl: To determine the pathological decline in intelligence in the HAWIE with the help of the breakdown quotient. In: Psychol. Forsch. 28, 1965, pp. 476-490.
  9. ↑ W. Baxa, E. Pakesch, S. Reitz: On the use of Dahl's short form of the Hamburg-Wechsler intelligence test. In: Wien.Z.Nervenheilkunde. 28, 1970, pp. 228-231.
  10. ^ W. Baxa, E. Pakesch: Communication on the use of an index at HAWIE to determine a secondary reduction in intelligence. In: Wien.Z.Nervenheilkunde. 30, 1972, pp. 119-130.
  11. ↑ W. Hartje, Berd Orgass: Proof of a HAWIE short form (WIP according to Dahl) in brain-damaged patients. In: Ztschr. F. experimental u. applied psychology. XIX, 1972, pp. 309-324.
  12. ↑ R. Gebhardt, H. Neumann: The intelligence as a function of the intelligence test carried out and of personality traits. A comparison of WIP and PMT. In: Psychother.med. Psychol. (Thieme). 26, 1976, pp. 23-31.
  13. ^ AI Rabin: A short form of Wechsler Bellevue Test. In: J. appl. Psychol. 27, 1943, p. 320.
  14. ↑ U. Seydel: HAWIE short forms and their cross validation. In: Diagnostica. (Hogrefe). 18, 1972, pp. 121-136.
  15. ↑ W. Demuth: Development of a HAWIE short form for examination and assessment. In: Swiss Arch. F. Neurology and psychiatry. 132, 1983, pp. 89-100.
  16. ↑ L. Thomas, G. Dahl: On the validity of intelligence diagnostics in the reduced changer test (WIP). In: Psychological contributions. (Verlag Anton Hain). XI, 1969, pp. 543-550.
  17. ↑ G. Dahl: To calculate the difficulty index with quantitatively graded task evaluation. In: Diagnostica. (Hogrefe). XVII, 1971, pp. 139-142.
  18. ↑ W. Baxa, E. Pakesch, S. Reitz: On the use of Dahl's short form of the Hamburg-Wechsler intelligence test. In: Wien.Z.Nervenheilkunde. 28, 1970, pp. 228-231.
  19. ↑ B. Orgass, W. Hartje: Proof of a HAWIE short form (WIP according to Dahl) in brain-damaged patients. II. Numerical agreement between WIP and HAWIE. In: Diagnostica. (Hogrefe). XX, 1974, pp. 22-30.
  20. ↑ W. Hamster: Empirical studies on the diagnostic valence of performance tests in clinical psychology. Dissertation. University of Tübingen, 1978.
  21. ↑ R. Olbrich: A control examination for a HAWIE short form (WIP von Dahl) in a heterogeneous clinical group. In: Ztsch. f. Clinical psychology and psychotherapy. 24, 1976, pp. 22-27.
  22. ↑ R. Schäuble, C. Gorlicki: Is a valid intelligence classification possible with a reduced form of the HAWIE-R? In: Diagnostica. (Hogrefe). 44, 1998, pp. 5-10.
  23. ↑ H.-J. Möller, G. Laux, H.-P. Kapfhammer: Psychiatry and Psychotherapy. Springer-Verlag, 2007, p. 434. (books.google.ch)
  24. ↑ B. Orgass, W. Hartje: Proof of a HAWIE short form (WIP according to Dahl) in brain-damaged patients. II. Numerical agreement between WIP and HAWIE. In: Diagnostica. (Hogrefe). XX, 22-30, 1974, p. 29.
  25. ↑ E. Brähler, H. Holling, D. Leutner, F. Petermann et al. (Eds.): Brickenkamp manual of psychological and educational tests. (Vols. 1-3). Hogrefe, Göttingen 2002.
  26. ↑ W. Demuth: How can the clinical usefulness of HAWIE general knowledge be improved for neuro-psychiatric patients? Attempt to re-validate. In: Fortschr. Psychiatr. (Thieme). 48, 1980, pp. 401-407.
  27. ↑ G. Dahl: To calculate the difficulty index with quantitatively graded task evaluation. In: Diagnostica. (Hogrefe). XVII, 1971, pp. 139-142.
  28. ↑ U. Tewes: HAWIE-R. Hamburg-Wechsler intelligence test for adults. Revision 1991. Verlag Hans Huber, Bern / Stuttgart / Toronto 1991.
  29. ^ Ingrid Baumert: Investigation of the diagnostic valence of the HAWIK and the development of a short form (WIPKI). Hans Huber, Bern 1973.
  30. ↑ Sigrid Hübner: Thinking and strategy behavior. Waxmann Verlag, 2000, p. 39 f. (books.google.ch)