Thursday, August 22, 2013

New proposed banding for PSLE

The current buzz among parents is of course, the recent National Day Rally announcements to upcoming changes to the school system, not least of all, the PSLE. I was asked to be interviewed for the Channel 5 programme Voices Today on this topic. At first I declined (it's on live tv! What could be more terrifying!) but then Mediacorp kindly said I could do the interview via phone, so I did.

Since I was going to do the interview, I had to organise my thoughts on the matter, so I thought I might as well share them here as well.

Why is the proposed banding system for PSLE good?

Currently, the way PSLE t-score is computed and released to the students is very unhealthy because it ranks each student in a linear form from first to last. If some children get the same t-score, the t-score then can move down to several decimal places to decide who gets placed before another.This sends the message that a child who gets say, 241 is somehow better than the one who got 240. Even worse, the one scoring 260.25 is better than another one who got 260.21.

As the PM said, this sort of fine stratification is meaningless. Because of the way secondary school admissions is done based on this t-score ranking, the competition is stifling. You try to outdo as many in your cohort as possible. You have students and parents clamouring for every last point, because losing out one spot in the ranking can mean not getting into your school of choice. It’s not about doing well, it’s about doing better than others.

I wouldn’t say this is the only reason for the tuition culture but it certainly exacerbates it. You know how we sometimes find it baffling that very bright students go for tuition in all subjects. But if you are aiming for the top school, eg RI/RGS/HCI/NYGS, you basically need to score 260 and above to be assured of a place (or maybe even more than that). Now, nobody, no matter how well they've been doing, would be cocky enough to think that they can guarantee that score. You're talking about the top 5% of cohort or so. So even if you’ve been scoring 95, you still go for tuition to try and chase that 100. Every bit matters. Whereas now, if scoring above 245 will mean you have a shot at the top schools, then there is less pressure. An A* is an A*.

On MOE’s part, the current system means that they have to keep accelerating the standard of the papers every year, or at least come up with novel questions that only the super bright will be able to work out because there is a need to differentiate the kids. This is another meaningless exercise because the purpose of the PSLE is to test what the kids have learnt, not test what they don’t know.

How should the banding be done?

I’m not sure how MOE will do the banding. One way is to just allocate points to the grades like O levels eg. A* = 1 pt, A = 2 pts, B = 3 pts, etc. However, since there are only four subjects, you probably would end up with too broad categories, ie too many kids falling within the same category of points.

To me, a fairer way is to keep the t-score formulation but release them in bands instead of absolute numbers. Eg. 245 and above (which I estimate to be about the top 15% of cohort) can be in one band. This way, the student or the secondary school will not know whether the student scored 245 or 260. The message we’re sending is that all these students are capable academically, there is no need for a finer distinction at age 12.

I’ve always felt that it’s easy for the top secondary schools to talk about their straight A graduates when they take in the top 5% of students to begin with. Most of these students will do well even if they were placed in a non-branded neighbourhood school. The more important question is, how much value did that top school give to the student? This new banding will be a truer test. If these top schools can take in a more diverse range of students and still produce the same number of straight A graduates, then it’s proof that the school helped to achieve the result.

Will this really lessen the pressure and competition?

It won’t eliminate competition totally because PSLE is still considered a high-stakes exam and students will still try to get into higher bands. However, it will definitely lessen the stress somewhat. For the very bright students who are already doing very well, there will now not be a need to chase the last point. That time can now be spent on more meaningful activities, like sports, arts, CCA, community service, etc.

In fact, I’m pretty sure that competition in these other areas will now increase because for the top schools, there will probably be more applicants within the band than spaces, in which case I'm not sure what will be the next step. Look at CCAs and other criteria? Do a separate GAT test? So you'll find that instead of going for tuition, to ensure a space, parents will send their children for extra sports/music lessons etc. And you'll end up chasing something else, like kids trained to be national swimmers from kindergarten. But if I were to look for the silver lining, at least it's less narrowly focused on academics.

At the end of the day, the system can only do so much. There will always be kiasu parents.

Why don't we just get rid of banding altogether and have pass/fail like in the past?

I think we need to acknowledge that some kids are just more able academically than others. We can't dispute that a 260 scorer is better at studies than a 200 scorer. Putting them together in the same class will not be an efficient model because they learn differently. But what banding signals is that there is no need to differentiate between a 245 scorer and a 260 scorer. They probably can learn equally well together. It's about creating the most optimal setting for learning. 


Anonymous said...

Dear Monica
So nice to read this. All the best for the interview!
A lot of parents are concerned if the T-Scores will still be released to the secondary schools despite the kids being placed in bands. This would mean schools can still pick and choose the kids based on their scores and it gives rise to anxiety. If parents and kids were given the T-scores, they may have a better sense of where they stand in terms of getting to a particular school.

monlim said...

Anon: This new banding will only work if schools do not know the exact scores either. Otherwise, it would defeat the purpose cos as you say, the schools will still pick the top scorers.

Ben said...

I am impressed by the depth of your insightful article.

I like your conclusion: "Why don't we just get rid of banding altogether". I think we should just make PSLE and O-level optional. As it is now, only those excel academically are allowed to opt out of O-Level (IP Program).

For O-level, if you talk to 100 students, more than 50 of them will tell you that they will go Poly. Most are very specific about what they want to do in Poly. Why do we have to ask these 50 or more to take 1 full year of Sec 4 to prepare for O-level? From past statistics, after sitting their O-level, we know that almost all 50 of them will still end up with Poly, as they so wish.

Singapore government is shifting educational policy towards "Every school a good school". This is just a first step. It will take us at least another 10 to 20 years to change to "Every student a good student". If the government is serious about this, they should consider the many ways of making PSLE and O-level optional. I am a bit disappointed that the education minister who is in charge of National Conversation is assigned with another huge task of Singapore 50th year celebration. He should instead move on the push for "National Conversation on education".

Anonymous said...

Mon, did i ever tell you this - in my first year of employment fresh from university, one girl in the same intake at the office was bragging about her scores (i.e. from prestigious schools all through... and 7 'o' levels or dunno how many points lah). One guy just turned around and said to her: "you 7 points, i 17 points. we end up in the same place."

we all burst out laughing. (and i married the guy 8 years later) :-)


monlim said...

KJJ: LOL!!!!! That's just classic!

Paul Ho said...

The only way is not to tweak the grading system. This broad banding of scores will truly be effective if there are more good schools around. There needs to a equitable shift of resources and manpower that will bring failing schools to the mid-band and the mid-band schools to the elite group.

One way is to designate fixed streams of students per school so that there would be a more equitable distribution of excelling students, good students and students who are attending the Normal Academic Stream.

IF the top ten schools can devote 3-4 classes to the banded group of top 10% of the N(A) students, then it will ensure a better spread of students to other schools, thus uplifting the schools that are struggling to succeed.

In order to prevent lower scoring students from taking advantage of the N(A) stream, MOE should make the judgement call on the cut-off between Express and N(A) students. Those who fall within the band of top 10% of the N(A) cohort gets to go to the elite schools.

At any rate, if these schools are really worth their salt, they should be able to turn these N(A) students to top scoring Sec 5 students at O Levels too.

monlim said...

Paul: There are too many assumptions in your suggestion, first that "good" schools are good because there are more resources and manpower. That's not true. Anyway, how do we define "good"? Currently, the top schools produce better results a big part of it is because they take in the academically bright students to begin with. Likewise, what you deem as "failing" schools, do worse because they take in less academically inclined students. I think the stats show that neighbourhood schools actually provide more value add, ie more kids graduate with better results than projected according to their PSLE t-score, than elite schools. I also know that MOE now assigns scholars and high achieving teachers to neighbourhood schools, not to elite schools.

I don't think taking in NA students will solve anything because these kids will still be grouped according to their results in their corresponding classes. It won't change the perception of how one point or decimal point in the t-score implies that one student is superior to another, which is the main reason for changing to the banding system.

kt said...

Dear Monica, thank you for this insightful write-up. I enjoyed reading it :-) Yes, I was also wondering why it will take so long for MOE to implement this change. One reason mentioned was for them to have sufficient time to raise the profile of more schools, so there will be more "good" schools around? I am looking forward to this as I think it will somewhat lessen the competition - there will be more schools with good niche programmes and run by dynamic heads of school.. I am worried as my child is in this "pioneer" batch and I certainly don't wish to have to experience more competition/stress :-(

Related Posts Plugin for WordPress, Blogger...