2012 PSLE Discussions and Strategy

[mommylow]

Chenonceau:

ruohoo97:

[quote=\"Chenonceau\"]The discrepancy in impressions is not surprising. Historically, PSLE Math has the widest spread in marks. There are very low marks and very high marks (i.e., the difference between the lowest mark and highest mark is large). In contrast, the language papers have the narrowest spread of marks (i.e., the difference between highest mark and lowest mark is smaller).


If you imagine all the PSLE scores plotted on a graph... there will be 4 bell curves looking like the following... See here for a picture of the bell curve - http://www.robertniles.com/stats/stdev.shtml

(1) Math - wide bell bottom
(2) Science - less wide bell bottom
(3) English - even less wide bell bottom
(4) Chinese - narrowest bell bottom

In order for EACH subject to have equal weightage, statistical operations have to equalize the spread across subjects. This means that they will have to have the same size bell bottom for all the subjects.

Now, if your bell bottom was narrow... you would have to pull it wider to fit the spread of the t-score curve. If your bell bottom were wider... you would have to compress the spread to fit the spread of the t-score curve. The t-score curves for each subject have EXACTLY the same spread.

This means that every 1 mark increase/decrease in raw score for Chinese (narrow curve that is pulled wider) gives you a higher increase/decrease in t-score than every extra mark in Math (that is compressed narrower) raw score.

A friend of mine works in After School Care... and there are some kids who could not finish HALF the paper. Yet, there are kids from Tao Nan, who complete with time to spare. This just follows the same pattern as every year's PSLE, there will be a wide spread in marks.

That is a very interesting point of view. I have always thought, if a paper is easy, T score will actually be lower, comparing to that of tough papers. Now you brought the spread of bell curve. how did you know the Maths curve is widest while Chinese curve is narrowest? Is there some publications by MOE? or KSPs? Thanks a lot.

Google the formula for t-score? You will find that it involves the standard deviation. Standard deviation is a measure of spread. Does this help?[/quote]

I learnt that if the general cohort does well, they will be bunched up towards the median score. The highest and lowest will also be nearer to the median score. In very difficult exams, the highest score will get higher and the lowest score even lower. That translates to easy papers give smaller standard deviation from the mean score.

[Chenonceau]

ruohoo... do these help?


http://www.greatminds.edu.sg/index.php?option=com_content&view=article&id=89:disclose-t-score-secret&catid=59:great-minds-club&Itemid=91


http://road-to-psle.blogspot.sg/2007/11/how-psle-aggregate-and-t-scores-are.html

The t-score formula has both cohort average (difficulty of paper) and cohort standard deviation (spread of marks around average). Both have a bearing on how much 1 mark raw score increase in a particular subject contributes to t-score. Most people look at paper difficulty... but given the formula, spread is important too.

http://i48.tinypic.com/245wldx.png\">

Read the above table with some pinches of salt... it is NOT confirmed that Chinese has the smallest spread of marks around the mean. If any parent can confirm/disconfirm this, it would be good. What the table above shows is ONLY that difficulty of paper (cohort average) and spread of marks (gap between best and worst student) has a bearing on t-score.

The relative spreads of Chinese, English, Science and Math are drawn from here - http://www.greatminds.edu.sg/index.php?option=com_content&view=article&id=89:disclose-t-score-secret&catid=59:great-minds-club&Itemid=91... this site reverse engineered the spreads from pass rate and quality pass rate for each subject.

[SAHM_TAN]

I think in the earlier posts, or maybe in another thread, there are a couple on PSLE that I’m a bit confused, parents were wondering abt the difficulty level of Eng and Maths papers. I view the posts by the students as response to those earlier posts, nothing more.


Best wishes to all P6 kids.

[fantasy1]

Hi


for the vocab, my DD chose decipher, but she said all her classmates had chosen determine

[ruohoo97]

Chenonceau:

ruohoo... do these help?


http://www.greatminds.edu.sg/index.php?option=com_content&view=article&id=89:disclose-t-score-secret&catid=59:great-minds-club&Itemid=91


http://road-to-psle.blogspot.sg/2007/11/how-psle-aggregate-and-t-scores-are.html

The t-score formula has both cohort average (difficulty of paper) and cohort standard deviation (spread of marks around average). Both have a bearing on how much 1 mark raw score increase in a particular subject contributes to t-score. Most people look at paper difficulty... but given the formula, spread is important too.

http://i48.tinypic.com/245wldx.png\">

Read the above table with some pinches of salt... it is NOT confirmed that Chinese has the smallest spread of marks around the mean. If any parent can confirm/disconfirm this, it would be good. What the table above shows is ONLY that difficulty of paper (cohort average) and spread of marks (gap between best and worst student) has a bearing on t-score.

The relative spreads of Chinese, English, Science and Math are drawn from here - http://www.greatminds.edu.sg/index.php?option=com_content&view=article&id=89:disclose-t-score-secret&catid=59:great-minds-club&Itemid=91... this site reverse engineered the spreads from pass rate and quality pass rate for each subject.

Thank a lot Chen, yes, greatly appreciated. It is indeed very hard to estimate Tscore based on test paper alone. For higher ability pupils, they may wish to have difficult paper to be a star scorer.

As for SD for Chinese and Maths, I am still bit confusing......

[Mathematica]

The paper was tricky, though not challenging. Some pupils said easy at first, only to realise their marks were lost after comparing ans with their classmates.

[GLORYmum]

Mathematica:

The paper was tricky, though not challenging. Some pupils said easy at first, only to realise their marks were lost after comparing ans with their classmates.

like your avatar very original.
some of dd's friend says quite easy but got 6 marks deducted.
maybe thought was easy at first so let guards down.

[mummy so kiasu]

Chenonceau:

ruohoo... do these help?


http://www.greatminds.edu.sg/index.php?option=com_content&view=article&id=89:disclose-t-score-secret&catid=59:great-minds-club&Itemid=91


http://road-to-psle.blogspot.sg/2007/11/how-psle-aggregate-and-t-scores-are.html

The t-score formula has both cohort average (difficulty of paper) and cohort standard deviation (spread of marks around average). Both have a bearing on how much 1 mark raw score increase in a particular subject contributes to t-score. Most people look at paper difficulty... but given the formula, spread is important too.

http://i48.tinypic.com/245wldx.png\">

Read the above table with some pinches of salt... it is NOT confirmed that Chinese has the smallest spread of marks around the mean. If any parent can confirm/disconfirm this, it would be good. What the table above shows is ONLY that difficulty of paper (cohort average) and spread of marks (gap between best and worst student) has a bearing on t-score.

The relative spreads of Chinese, English, Science and Math are drawn from here - http://www.greatminds.edu.sg/index.php?option=com_content&view=article&id=89:disclose-t-score-secret&catid=59:great-minds-club&Itemid=91... this site reverse engineered the spreads from pass rate and quality pass rate for each subject.

Smaller spread of mark from the mean will lead to higher or lower T score?

[alng]

mummy so kiasu:

Chenonceau:

ruohoo... do these help?


http://www.greatminds.edu.sg/index.php?option=com_content&view=article&id=89:disclose-t-score-secret&catid=59:great-minds-club&Itemid=91


http://road-to-psle.blogspot.sg/2007/11/how-psle-aggregate-and-t-scores-are.html

The t-score formula has both cohort average (difficulty of paper) and cohort standard deviation (spread of marks around average). Both have a bearing on how much 1 mark raw score increase in a particular subject contributes to t-score. Most people look at paper difficulty... but given the formula, spread is important too.

http://i48.tinypic.com/245wldx.png\">

Read the above table with some pinches of salt... it is NOT confirmed that Chinese has the smallest spread of marks around the mean. If any parent can confirm/disconfirm this, it would be good. What the table above shows is ONLY that difficulty of paper (cohort average) and spread of marks (gap between best and worst student) has a bearing on t-score.

The relative spreads of Chinese, English, Science and Math are drawn from here - http://www.greatminds.edu.sg/index.php?option=com_content&view=article&id=89:disclose-t-score-secret&catid=59:great-minds-club&Itemid=91... this site reverse engineered the spreads from pass rate and quality pass rate for each subject.

Smaller spread of mark from the mean will lead to higher or lower T score?

Higher mean and higher SD will lead to lower T-score.

[Chenonceau]

mummy so kiasu:

Smaller spread of mark from the mean will lead to higher or lower T score?

Hello dearie... smaller spread higher contribution to t-score per 1 mark. Wider spread, smaller per 1 mark. Sorry I can't explain very well... not good at Math.