Comments on: Is the uni admissions system in ‘crisis’? http://andrewnorton.info/2007/01/is-the-uni-admissions-system-in-crisis/ Observations from Carlton's Lone Classical Liberal Fri, 09 Jan 2009 22:48:14 +0000 http://wordpress.org/?v=2.7 hourly 1 By: Club Troppo » Wednesday’s Missing Link http://andrewnorton.info/2007/01/is-the-uni-admissions-system-in-crisis/comment-page-1/#comment-4047 Club Troppo » Wednesday’s Missing Link Wed, 24 Jan 2007 04:33:36 +0000 http://andrewnorton.info/blog/2007/01/19/is-the-uni-admissions-system-in-crisis/#comment-4047 [...] Is the uni admission system in ‘crisis’? - No, argues Andrew Norton. [...] [...] Is the uni admission system in ‘crisis’? - No, argues Andrew Norton. [...]

]]> By: Damien Eldridge http://andrewnorton.info/2007/01/is-the-uni-admissions-system-in-crisis/comment-page-1/#comment-4046 Damien Eldridge Fri, 19 Jan 2007 06:34:32 +0000 http://andrewnorton.info/blog/2007/01/19/is-the-uni-admissions-system-in-crisis/#comment-4046 I imagine that it is certainly true that if you reduce the variation in the sample, then you will find it harder to get significant correlation between school marks and uni marks. But why would you want to do this? If people with very high school marks do very well at uni and people with very low marks do very poorly at uni, then this tells you something useful. It is not caused by a miscoding of the data or some irrelevant extraneous factor, which would presumably be the basis for getting rid of an outlier. I imagine that it is certainly true that if you reduce the variation in the sample, then you will find it harder to get significant correlation between school marks and uni marks. But why would you want to do this? If people with very high school marks do very well at uni and people with very low marks do very poorly at uni, then this tells you something useful. It is not caused by a miscoding of the data or some irrelevant extraneous factor, which would presumably be the basis for getting rid of an outlier.

]]> By: conrad http://andrewnorton.info/2007/01/is-the-uni-admissions-system-in-crisis/comment-page-1/#comment-4045 conrad Fri, 19 Jan 2007 01:42:01 +0000 http://andrewnorton.info/blog/2007/01/19/is-the-uni-admissions-system-in-crisis/#comment-4045 Actually, I think the correlation in many -- perhaps most -- courses based on high school results is likely to be even less one you get rid of the left and right outliers because a) It is hard to report non-significant results in many areas. Thus there are probably piles of .1 and .2 correlations floating around. b) No university is going to want to say that their admissions procedure doesn't predict performance at all. Actually, I think the correlation in many — perhaps most — courses based on high school results is likely to be even less one you get rid of the left and right outliers because

a) It is hard to report non-significant results in many areas. Thus there are probably piles of .1 and .2 correlations floating around.

b) No university is going to want to say that their admissions procedure doesn’t predict performance at all.

]]>