Page 159 - DEDU504_EDUCATIONAL_MEASUREMENT_AND_EVALUATION_ENGLISH

Unit 11 : Development of Norms of a Test

(c) These norms are useful in all types of circumstances, such as educational, industrial, Notes
military fields etc.
(d) Percentile norms can be easily developed.
(e) They can be used to meaningfully express the scores with different units and numerical
standards.
(f) Generally only percentile norms are ascertained for personality tests, IQ tests, attitude
tests, aptitude tests etc.
Limitations
(a) It is not possible to carry out statistical analysis of these norms.
(b) The percentile scores of different tests cannot be compared unless the groups on which
they were administered are not comparable; for example, if in a personality test, percentile
norms have been developed for adolescent girls taken from a large group, then the scores
of all adolescent girls can be compared with these.
(c) In normal situations, percentile norms tell the relative position of each individual, but it
does not make out the difference in scores between two individuals.
(d) Percentile norms are often confused with percent scores.
(e) The relative position of an individual is ascertained on the basis of these norms. It is not
possible to analyze actual ability or capability of an individual objectively.
(f) The units of percentile scores are not uniform. If the details of actual scores are almost
common, then there is much difference in changing proximate scores into percentile
values, while there is not much difference in changing scores at extreme ends.
4. Standard Score Norms : The greatest shortcoming of percentile norms is that the units of
scores is not equal in this, that is, the two consecutive percentiles are not equally or uniformly
distanced. For example, the difference between 30th and 40th percentiles is not equal to the
difference between 60th and 70th percentiles. Due to this shortcoming, these norms cannot
be used to compare the differences among different candidates. Therefore, test-makers look
for such units which are meaningful throughout the entire expanse. From this standpoint,
the standard score norms are widely used. These norms are also called Z score norms.
By standard score norms is meant to change the raw scores of candidates into standard
scores. This type of norms are found out with the help of standard deviation (S.D. or σ ).
This standard deviation is a measurement of the expanse of scores of a group. Standard
norms are based on normative group. These norms analyze the achievement of an individual
on the basis of his scores in the context of the particular group. Because these express
uniform units, so they are different from percentile norms. Their basic unit is the standard
deviation of the reference group, and the scores of an individual are expressed below
median of the group and above standard deviation unit.
Basically, all these are the same thing, and are the modified forms of the same scale.
The mean and standard deviation of Z-score are 0 and 1 respectively. The mean and standard
deviation of t-score are 50 and 10 respectively. The mean and standard deviation of Sten-
score are 5.5 and 4 respectively, and those of Stanine-score are 5 and 2 respectively. T-score
was first used by McCall and Sten-score was first used by R.V. Cattell. If in a test, the mean
of a large group is M and standard deviation is σ , then the scores can be calculated by the
following scores :
Z – Score = X – M σ
(
t – Score = 50 + X – M σ )×10
Sten – Score = 5.5 = (X – M σ )× 2

Stanine – Score = 5 = (X – M σ )× 2

LOVELY PROFESSIONAL UNIVERSITY 153