Page 179 - DEDU504_EDUCATIONAL_MEASUREMENT_AND_EVALUATION_ENGLISH

Unit 13 : Interpretation of Test Scores : Qualitative and Quantitative

As in any such table, the limits of each category were arbitrarily determined. These and similar Notes
categories are intended to serve only as guides in the interpretation of intelligence quotients and
for purposes of statistical classification and analysis.
There are a number of problems associated with the interpretation and use of the IQ that are
explained in later sections of this and other chapters. Our purpose at this stage is primarily to
define and explain the meaning of the concept.

The use of this index was first suggested by Stern (17) and Kuhlmann (12) in 1912;
but it was not actually employed as part of test findings and reports until 1916 when
the first edition of the Stanford-Binet scale was made available.

13.1.5 Deviation IQ
The “deviation IQ” is an adaptation of the standard score (z) technique. The method of determining
the deviation IQ can be shown by using the Wechsler test’s procedure as an illustration. For each
individual, the raw score is converted into a weighted score by using a conversion table. The
mean weighted score of the group is given a deviation IQ value of 100; the standard deviation of
the scores is equated with a deviation IQ value of 15. Thus, a person whose point score places him
at — 1 SD will have a deviation IQ of 85. One whose score is at —2 SD will have a rating of 70.
Similarly, positive SD values will give ranks above 100 : +1 SD equals 115 deviation IQ; +2 SD
equals 130; and so forth.
The 1960 revision of the Stanford-Binet scale also uses the deviation IQ, calculated by a different
method; but the basic principle is the same. The principle is that an individual’s intelligence
quotient should be determined by the relative extent to which his score on the test deviates from
the mean of his age group, and that an intelligence quotient of a given value should have the
same relative significance throughout the age range. These ends are now achieved by using units
of standard deviation as the basis; hence the name of the new index.
Making the mean score equal to a deviation IQ of 100 is readily understandable, since this value
has long been conventional and is accepted as representing the average or normal. It also appears
that the most probable standard deviation of intelligence quotients is 15-16, as found with the
Stanford-Binet (which in many ways is regarded as a criterion); hence 15 has been taken to
represent the standard deviation of the newer index. The choice of this value, therefore, was not
an arbitrary one. Furthermore, the distribution yielded by a standard deviation of 15 points is
very similar to the one to which psychologists and educators have become accustomed, and in
which values at each of the several levels have acquired qualitative significance in regard to
mental ability and educational promise.
The deviation IQ, furthermore, is especially useful at age levels above 16 or 18 years. For these
and older persons, the use of mental age and the formula for the ratio IQ (MA/CA) have been
regarded as inappropriate and questionable by many psychologists. (This aspect of mental age is
discussed in Chapter 10.)
It should be clear, from the materials thus far presented, that percentiles, standard deviations,
standard scores, and intelligence quotients are intimately related. Whatever index is used, its
principal significance is found in the relative rank it represents and in its psychological, educational,
and vocational connotations.
Although the primary purpose of this section is to define and explain these concepts used in
psychological testing, it is relevant here to emphasize the qualitative aspects of these indexes.

LOVELY PROFESSIONAL UNIVERSITY 173