PSLE 2009 - Chinese Paper

[concerned]

the chi compo is about the girl who thows away food that her mom prepared but decided to give it to a boy who did not hav money to buy food during recess.the other topic is 'this caused me not to be rude again.tis is according to my daughter.dunno if correct. :idea:

[janice_koh]

today my son forget to bring wallet for recess....maybe i should tease him if his class prefect pass him any lunch boxes.

[sunshinemum2]

CrazyDad:

kids say this paper is very easy. If this is the case, it will be like normal, 2% of cohort <50>75 mark. This will probably contribute to standard deviation of <10, about 9 maybe and average of 78. You can estimate your score as followed;

Score\tAverage\tSTD DEV\tT Score
100\t78\t9\t74
85\t78\t9\t58
75\t78\t9\t47

Do you have estimate for other subjects (English and Maths)based on their level of difficulty this year?

[Dharma]

turquoise:

ApronMama:

[quote=\"blitz\"]My gosh - so one might pass the written exam, but the T-score will reflect as failed?


Score Average STD DEV T Score
75 78 9 47

If paper easy, most people score well, T-score will be low. No good for those are really good in Chinese. Guess higher Chinese will be difficult.

What does it mean for the average scorers then?[/quote]Hi turquoise,

The T-score is calculated based on the following formula

T- Score = 50 + 10((X – Y)/Z)

Where

X = child’s raw score
Y = average score of the whole cohort
Z = standard deviation

Z = Squareroot( Summation of(X – Y)2/N)
X = raw score of each child, Y = average score of cohort and N = population size

The estimate T-score of 47 is obtained as follows

T-score\t = 50 + 10 (75-78 )/9
\t\t= 46.7 (rounded to 47)

If the child’s raw score is about the average score, then the T-score will be round about 50.

[turquoise]

Thanks, Dharma. Am I correct to assume that generally, if a paper is easy, there will be a low t-score and that is not in the favour of those good in the subject.


Conversely, if the paper is very difficult with the general population performing poorly, then the ones who score very well will have a high T score?

Thanks.

[Dharma]

turquoise:

Thanks, Dharma. Am I correct to assume that generally, if a paper is easy, there will be a low t-score and that is not in the favour of those good in the subject.


Conversely, if the paper is very difficult with the general population performing poorly, then the ones who score very well will have a high T score?

Thanks.

Hi turquoise,

If the paper is easy, the average score will be pushed up. If most pupils have scored around the average, the variance will be small and therefore the standard deviation will be low. With a low standard deviation, although the average mark is high, a pupil who is strong at the subject will have high T-score.

However if the paper is difficult, the average scores will be lower but there is likely to be greater spread of the marks around the average score. We are likely to get a greater standard deviation here. Here, a strong pupil will still have a high T-score as the higher standard deviation will be offset by the higher difference of his score as compared to the average score.

What I think the difference will be is that for every mark drop in the raw score will result in a bigger drop in the T-score when you have a simpler paper as compared to a more difficult paper.

[von72]

Dharma:

Hi turquoise,


If the paper is easy, the average score will be pushed up. If most pupils have scored around the average, the variance will be small and therefore the standard deviation will be low. With a low standard deviation, although the average mark is high, a pupil who is strong at the subject will have high T-score.

However if the paper is difficult, the average scores will be lower but there is likely to be greater spread of the marks around the average score. We are likely to get a greater standard deviation here. Here, a strong pupil will still have a high T-score as the higher standard deviation will be offset by the higher difference of his score as compared to the average score.

What I think the difference will be is that for every mark drop in the raw score will result in a bigger drop in the T-score when you have a simpler paper as compared to a more difficult paper.

Hi Dharma, thanks so much for sharing!!

[finder]

what the typical range for the standard deviation ? Tks.

[Dharma]

finder:

what the typical range for the standard deviation ? Tks.

Sorry I am not sure but can actually work it out using hypothetical figures on a spread sheet by imputing estimated raw of pupils scores based on a certain sample size. You can calculate the average score and SD. For difficult papers the average may be in the range of 50 to 60 but simpler papers can estimate about 70 to 80.

My guess is that SD normally will be lower, in the range of 10, possibly for difficult papers and probably in the range of 15 to 22 for tougher ones. However, if you are able work out using excel and that will help to calculate the SD and T-scores.

Mind you, what you calculate is just as good as your guess…. Too many uncertainties to be sure.

[mathsparks]

Interested in a copy of the excel T-score calculator? Pls see


http://www.kiasuparents.com/kiasu/forum/viewtopic.php?t=5858&start=0&postdays=0&postorder=asc&highlight=