2022 PSLE Discussions & Strategies (born in 2010)

[SG_KP1]

Technospaz\" post_id=\"2074889\" time=\"1658039985\" user_id=\"135989:

There's no bell curve? Really?

Yes, that is correct. MOE has stated several times that the AL scores awarded are strictly based on raw marks.

The distribution of scores likely isn't normal (see my last comment), but is somewhat normal, and in every year there will most likely be a reasonably distribution of scores. But the number of students that score AL4, AL5, AL6, etc is not fixed and can go up or down based on cohort performance/paper difficulty.

[sembgal]

Technospaz\" post_id=\"2074889\" time=\"1658039985\" user_id=\"135989:

There's no bell curve? Really?

No bell curve for the AL score for PSLE. There was bell curve previously for T-score.

Whether there’s bell curve for the marks awarded for Oral exam will remain a mystery.

[sembgal]

SG_KP1\" post_id=\"2074897\" time=\"1658041116\" user_id=\"188234:

They've never clarified whether 50 percent of students above/below 75 applies for all subjects the same. But yes, generally it would seem that AL16 is about the median score.

If you look at the old t-score system, the average, or more specifically the mean, subject score got you 50 subject t-score points. I.e. if you get the average score for all four subject you would get a t-score of 200. However, based on the percentage of students that qualified for Express (something like 66% or so), it could be inferred that the median subject score > mean subject score.

I think this is also likely to be true because the subject score is capped at 100 (+25 vs. 75) but can fall all the way to zero (-75 vs. 75). Also, 75 x 4 = 300 raw point out of 400. Rough t-score estimate would be ~220 (0.735 multiplier). I.e. median t-score > 200.

AL 16 = T-score 200 ?

[SG_KP1]

sembgal\" post_id=\"2074899\" time=\"1658042678\" user_id=\"5559:

No bell curve for the AL score for PSLE. There was bell curve previously for T-score.

Whether there’s bell curve for the marks awarded for Oral exam will remain a mystery.

There was no official bell curve for t-score. T-score was essentially a z-score and in any normal or semi-normal distribution there will only be a certain number of z-scores = 1, 2, etc. But I don't believe there was a \"forced\" distribution each year that fixed the number of 260, 270, 280, etc t-scores (260 = z-score 1.5 for each subject, 280 = z-score = 2.0 for each subject).

There is a debate over whether the subject grades of A*, etc were fixed at 91 or floated based on cohort performance.

[sembgal]

https://www.moe.gov.sg/news/press-releases/20220228-director-general-of-education-retires-after-38-years-of-service


New leadership takes over

[SG_KP1]

sembgal\" post_id=\"2074900\" time=\"1658042792\" user_id=\"5559:

AL 16 = T-score 200 ?

Likely no, it's the difference between median and mean. See my prior post.

[sembgal]

SG_KP1\" post_id=\"2074903\" time=\"1658043027\" user_id=\"188234:

Likely no, it's the difference between median and mean. See my prior post.

Thanks for your clarification.

[SG_KP1]

sembgal\" post_id=\"2074904\" time=\"1658043156\" user_id=\"5559:

Thanks for your clarification.

No problem. Another rough data point/check is:

SNGS had affiliate t-score of 230, new AMR is AL14
MGS/SCGS had affiliate t-score of 220, new AMR is AL17
Various CHIJ had affiliate t-score of 200, new AMR is AL20

Not saying there is a direction translation between the two scoring systems (there isn't). But if you take the schools'/MOE's perspective of setting the bar roughly equivalent so that a similar number of students qualify (or hit a similar standard) the figures above can give some insight or sanity check.

[zac's mum]

sembgal\" post_id=\"2074900\" time=\"1658042792\" user_id=\"5559:

AL 16 = T-score 200 ?

T-score 200 = AL20. At least judging from:
https://i.imgur.com/padYpld.jpg\">

[sembgal]

zac's mum\" post_id=\"2074912\" time=\"1658044158\" user_id=\"53606:[quote=\"zac's mum\" post_id=2074912 time=1658044158 user_id=53606]
T-score 200 = AL20. At least judging from:
https://i.imgur.com/padYpld.jpg\">[/quote]

Hi

Interesting data. Where is the data source from ?