RE-THINK SCHOOL RANKING

	Ranking schools according to their Mean Subject Grade (MSG) is meant to
help parents make informed choices when selecting a school. The Education
Minister recently said that only the results of English and the five best
subjects would henceforth be used to calculate the MSG. I would like to
suggest two more refinements: using the Median Subject Grade; and improving
the performance classification system.
	Firstly, ranking schools on a relative scale is conceptually flawed. Since
the objective is informing parents how well a school prepares its students
for the national examinations, performance should be measured against
objective examination results rather than by relative ranks derived from
small differences in the MSG.
	The reliance on the MSG as the indicator of school performance is
deplorable. The MSG is misleading because poorer grades have greater
weight: an F9 has the weight of nine A1s. Therefore, schools whose students
generally do well may yet have relatively big MSGs because of a few
failures. The greater the variation in student performance, the more
unreliable the MSG. Worse, the simplistic bundling of grades from different
subjects denies parents relevant information on school performance in the
various subjects.
	A more informative gauge of school performance is the Median Subject
Grade, which is the grade obtained or bettered by 50 per cent of the
students in a subject. Percentile-based indicators remove the weight bias
in the MSG. We could be stricter and use the 75th Percentile Subject Grade,
which is the grade obtained or bettered by 75 per cent of the students. We
could even report the subject grades at the 25th, 50th and 75th percentiles
to reveal the performance distribution.
	Reporting the Median Subject Grade for each subject allows parents to
judge the strength of a school in each subject. They can pay more attention
to those subjects that their children intend to take and discount the rest.
They will also know the probability of their children obtaining a
particular grade in each subject.
	The second refinement is to replace the present system of classifying
schools as Top Ten or Top Fifty on the basis of the MSG. Instead, we should
classify schools as "A-average", "B-average", "C-average" or "Unclassified"
in each subject on the basis of their Median Subject Grades. "A-average"
schools would have Median Subject Grades of A1 or A2, "B-average" schools
Median Subject Grades of B3 or B4, and "C-average" schools Median Subject
Grades of C5 or C6. Schools with Median Subject Grades below C6 would be
"Unclassified".
	In the proposed system, any number of schools can be "A-average". This
relieves the unrelenting pressure on teachers and students to improve their
school"s relative rank regardless of how well the school is doing. It also
corrects the schools" diversion of ever more resources to ranking at the
expense of other educational objectives.
	Of course, the compilation of such information is more burdensome than
calculating the MSG; but any information the authorities provide should be
quality information.



Yours sincerely

 
_____________________
Chong Fu Shin Francis