The Editor, Sir:

I write in response to an article by Rev Devon Dick titled 'GSAT Gestapo scoring system', published July 8, in which, in regard to the GSAT scoring system, he comes to the conclusion: "So, it does not matter whether the Ministry of Education uses standard scores or percentages, it should not affect who is classified as the top performer." .

It does matter whether standard scores or percentages or used, and I think it is important that this matter be properly understood, because it affects not only the classification of the top performer but the ranking of all students.

I will attempt to bring clarity to this matter without getting too technical.

The only situation in which it definitely does not matter is a case where a student has the highest percentage in each subject, because he will, therefore, also have the highest standard score in each subject.

Standard score

On the issue of raw scores, percentages, and standard scores, if a raw score and its corresponding percentage are converted to a standard score, they will both produce the same standard score.

The standard score takes into account the performance of all students in the group and, essentially, determines how far each student is, above or below the average mark, regardless of what the actual average mark is.

The distance is measured not in percentage points but in standard deviations. A standard scoring system will show, for example, that an athlete, who is one second slower than Usain Bolt's 100m world record of 9.72 seconds, is much further from breaking this record than an athlete who is one second slower than the 5000m world record of 12 min 37seconds.

I do agree that the concept of a standard deviation could be considered Gestapo-ish, and, I quote, "a secret and only a chosen few know it". However, the concept of the end result, the standard score, is fairly simple and I will illustrate with a simple example.

If, for a particular group of students, the average score for mathematics was 80 per cent, and the average score for science was 90 per cent, a student scoring 81 per cent in mathematics would have a higher standard score than a student scoring 90 per cent in science for the simple reason that 81 per cent in mathematics was above average while 90 per cent in science was just average.

The beauty of the standard scoring system is that a student, for eample, who is excellent in mathematics, and is counting on his performance in mathematics to boost his overall average, will not be affected much if, for his particular GSAT year, the math test is unusually difficult.

As long as his performance is excellent, compared to other students, his standard score will also be excellent regardless of whether his percentage score is 70 per cent or 95 per cent.

Other pass marks

For the purpose of ranking GSAT students, a standard scoring system is fair and far superior to just using percentages. The reason this system is not used more often is because calculating standard scores is far more tedious and technical than just calculating a percentage and not as easy to interpret.

Also, in a properly functioning school system, there are important pass marks and standards. So, for example, if scoring below 60 per cent on a math test, is deemed unacceptable, then the fact that the top performing student scored 55 per cent and, therefore, had the highest standard score does not change the fact that the performance is unacceptable.

Using percentiles

To prevent further disputes, the Ministry of Education should provide percentiles for each student, showing the percentage of students that performed better than for each subject.

Percentiles are easier to interpret and are a better reflection of a student's ranking.

It is important that the GSAT selection process is perceived to be fair and transparent, and that students and parents are aware of exactly how students are ranked and, therefore, placed.

I am, etc.,

SCOTT ROBINSON

scottroninson@hotmail.com