Another dot in the blogosphere?

#PSLE2021 (Part 3): Differentiation vs granularity

Posted on: July 22, 2016

In Part 1 of my analysis of the new PSLE assessment system, I highlighted the important fundamental switch from norm-referenced testing to criterion-referenced testing.

In Part 2, I described how the summative design and implementation of the high-stakes exam counters positive change.

In this part, I reflect on MOE’s and the public’s obsession with differentiation when we all should be more concerned about granularity.

The current PSLE uses transformed scores (T-scores) and their aggregates as outcome measures for the exam. There are at least three major problems with doing this.

  • The scores can be normalised (see Part 1) and this process is not transparent to the students or the public.
  • A score might indicate where a student stands relative to his or her peers, but it does not indicate what that number means. The student does not know what areas of learning need to be addressed because the exam papers are not returned and there is no feedback loop. This is typical of summative assessment (see Part 2).
  • The aggregate scores seem to finely differentiate students who are competing for places in Secondary schools. For example, a student with an aggregate score of 221 gets in, but another with 220 does not. Someone sets an entry score, but no one can really explain why that benchmark is and what it means. That is one way we play the cruel numbers game.

PSLE2021 is supposed to take away that differentiation because exam papers are graded with Achievement Levels (ALs). Students are assessed on four academic subjects and each subject can be graded from AL1 to AL8. This results in 29 discrete categories of aggregate AL scores of 4 to 32. There is a better, but still not adequate, granularity of scores.

It is no surprise that other bloggers and their mothers have pointed out that the competition will now be for low aggregate scores. Tuition centres might tweak their marketing material to focus on lowering AL scores.

The schooling arms race used to be to get the highest aggregate T-score possible. The next battle is to get the lowest aggregate AL score possible. This is like a Pokémon Go game, just not as fun. You do not want to collect all the scores. You only want the very rare AL1 Pokémon.

The “new” PSLE does not change or break the summative assessment mould. It is still a sorting tool. With the ordered choice of Secondary schools by students playing a more important role than the current model, it is being tweaked to be a decision-making tool.

If the focus is on student achievement and learning — as claimed by the MOE PSLE2021 microsite and echoed the press — students need even greater granularity. By this I am not referring to exact scores of each paper, although this will provide more coarse insight. I am also referring to feedback and remediation on areas of weakness.

At the risk of painting a overly simplistic dichotomy, we have these divergent paths:

  1. Summative assessment model, T-scores, fine differentiation for the purpose of sorting.
  2. Alternative assessment model, high granularity for the purpose of meaningful learning.

Our MOE seem to be designing a hybrid, at least on paper. In its wish list is: Summative assessment, some granularity, a focus on learning. This is a very elusive Pokémon, it is exists.

What paths will we take? What game will we play? What is at stake?

1 Response to "#PSLE2021 (Part 3): Differentiation vs granularity"

CHAN Hsiao-yun 曾曉韻: RT @ashley: #PSLE2021 (Part 3): Differentiation vs granularity #edsg via


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Click to see all the nominees!

QR code

Get a mobile QR code app to figure out what this means!

My tweets


Usage policy

%d bloggers like this: