Page 122 - DEDU504_EDUCATIONAL_MEASUREMENT_AND_EVALUATION_ENGLISH

Educational Measurement and Evaluation

Notes though average chronological age of a 7th grade student may be 13 years and 6 months at the
end of the school year, the average test score of pupils of 13 years and 6 months is not at all the
same as the end of the year score for the 7th grade. Actual achievement of underage pupils is
significantly superior to that of overage pupils in a given grade. For developers of standardised
tests therefore the need for norms that take into account the wide differences in maturity, mental
ability or school progress within the grade cannot be undermined, to avoid interpretations that
are likely to be misleading.

9.3.4 Age-at-grade Norms
In the process of test standardisation, establishment of age-at-grade norms involves difficulties
like : (a) availability of population, and (b) statistical procedures resulting from inadequate
population groups. For example, the number of 6th grade pupils who are 10 to 11 years old
would represent a large portion of normal 6th grade population, whereas the number of underage
pupils in 6th grade who would be 9, 8 or 7 years old and the number of overage pupils who
would be 11, 12,13 or 14 years old would fall off rapidly. Therefore in order to get reliable age-
at-grade norms for all ages within the grade, very large number of pupils have to be tested to
secure adequate population in the fringe areas, failing which estimation by extrapolation has to
be made. In the IOWA Basic Skill Test-C, language raw point scores are directly converted into
grade-equivalents as shown below :
6th grade : 1st Semester, 13 yr and 6 months old pupil has grade-equivalent score of 4.8
6th grade : 1st Semester, 10 yr and 6 months old pupil has grade-equivalent score of 6.3
6th grade : 2nd Semester, 13 yr and 6 months old pupil has grade-equivalent score of 5.3
(Typical Child) 9th grade : 2nd Semester, 13 yr and 6 months old pupil has grade-equivalent
score of 10.1.
It is apparent that age-at-grade norms offer very useful means for interpretation of individual
accomplishment, especially for pupils who are over-age or under-age for that grade.

It is interesting to note that younger pupils within the grade make higher grade-
equivalent scores up to a certain point and the older pupils make lower.

9.3.5 Percentile Grade or Subject Norms

This is done by turning raw scores or standard scores into percentile scores by computing
percentile values from the frequency tables for each grade and assigning percentile equivalents
for each score. Percentile norm labels show a wide sampling of pupils in a certain grade or
course : (a) the percentage of pupils exceeding each score or each of a number of equally special
scores, or (b) the score below which certain percentages of pupils fall. Percentile scores
corresponding to specific raw or standard scores may be reported by grades by test parts and
totals, or only the raw scores or standard score equivalents for specified percentiles, quartiles
and deciles may be shown in more compact tables. It is suggested that when percentile grade
norms are provided, these values should be used in the interpretation of individual pupil scores
rather than the grade equivalents from usual grade norms. Overlapping of the distribution of
scores from grade to grade on most of the standardised tests at elementary level is quite great so
that the differences between successive grade medians are so slight that grade equivalents may
exaggerate differences, leading to unsound interpretation. A 7th grade pupil assigned a grade
equivalent of 9.5 does not belong to 9th grade. It is more accurate to use percentile scores to
describe a pupil’s achievement as superior in relation to other 6th grade pupils.

116 LOVELY PROFESSIONAL UNIVERSITY