Should low ATAR students be admitted to university?

Over at Catallaxy, Judy Sloan is having a go at low ATAR university courses.

I just want to have the bridges identified which are designed by civil engineers with cut-off points of 62.

And I also noticed that the cut off score for entry into Primary Education courses is in the 50s – pity the poor children in a few years time.

As Judy hints at, ATAR (or its predecessors: ENTER, UAI, TER) is only moderately predictive of future academic performance, and even then only for higher ATAR students. This overview paper on Victorian university selection practices summarised some of the research:

Their … work at Monash confirmed the correlation between high ENTER and strong university performance (r=0.38 for ENTER over 80). Importantly, however, they found little correlation between ENTER and university performance for low to middle ENTER bands (r=0.04 for ENTER below 80). This finding supports that of Murphy et al. (2001), who found in their study of RMIT students that the strongest correlations between ENTER and university performance were at ENTERs above 80, with no correlation between 40 and 80 and variable correlation below 40.

However completions data published in the base funding review final report suggests that ATAR might be more predictive of completion than grades. Less than half of students who entered university in 2005 on an ATAR between 30 and 59 had completed by 2010, compared to over 80% of the 90+ ATAR students. (This is the best completions data published in Australia, as with the national student number system started in 2005 we can finally track the many students who change universities.)

The issue raised by low ATAR students (whose numbers are growing) is as much about whether their own interests are being served as whether incompetent professionals might come out at the other end. If assessment systems are robust, they can be failed if they don’t meet the standards. And they will be sacked if they are passed but don’t measure up in the workplace.

But the student may have wasted their time, money and effort on study that doesn’t go anywhere. The tiny correlations between ATAR and grades suggest that this screening device should not be used to prevent applicants enrolling. But they should be properly informed of the non-completion risks. And we need a lot more research to see if there are other factors that allow us to better predict who will succeed and who will not.

  1. I must say I’ve got data from some of the courses where I work that doesn’t show this pattern. Even in my own course, we have two campuses that teach exactly the same thing, but the ENTER has historically been about 25 points different. The overall average difference has never been less than about 10% on any piece of assessment ever, and there is a qualitative difference too — we have to try extremely hard just to get the people in the campus with the low ENTERs to pass, let alone get good marks. My bet is that you find the same discrepancy at, for example, Monash Clayton vs. Gippsland, Latrobe vs. Latrobe Bendigo etc. . One guy is also extremely good about collecting this sort of data (I think he has more than a decades worth), and there are even differences in correlations within subjects — some do correlate and some don’t.

  2. Judy is not to good at history but most of the engineers used in the Snowy Mountains only went to what is now TAFE! There were few, if any that were Uni graduates.

    Any book on John Monash would give you a useful guide to building bridges as well.

  3. okay, I got around to reading that report. I think this paragraph says it all:

    “Dobson and Skuja also found that correlations differed with area of study. ENTERs were
    most predictive of first year performance in engineering, agriculture and science (r=0.59 to
    0.61) and least predictive in management courses (r=0.33), education (r=0.11) and health
    (r=0.10). The authors suggest that ENTER’s predictive capacity diminishes as subjects move
    away from areas taught at school (Dobson & Skuja 2005: 55).”

    You can make your own mind up as to why the last three types of course have no great correlations with ENTER scores, and whilst I agree with with the last paragraph I also think it’s probably also contingent on what is being taught. At least for education, looking at some of the bizarro suggestions from teachers that have been generated from the silly proposal by UA about Aboriginal content, and the fact they all use terms from post-modern sociology (“lived perspective”,”knowledges”…other irregular plurals that are incorrectly pluralized etc.), perhaps you really do need a different mindset to do well in these courses.

  4. Conrad the correlation by course may suggest that it matters where the knowledge from the pre-requisite subjects really are important.

    Lots of science and engineering subjects rely on you already knowing calculus, matrices and trig. Students entering the physical sciences with high ATARs presumably did well at high school maths and therefore have quite a head start on the university content.

    However in lots of other areas that specific high school knowledge isn’t as useful.

  5. Andrew I think the unfixable issue is that ATAR scores are terrible at measuring anything other than being a single sample of a students ability to jump through a particular set of hoops and playing the game.

    Anecdotally I know of many younger siblings who benefit vastly from their families learning process with the first kid to pass the year 12 gauntlet.

  6. M- There are study skills and practices that would help get a higher ATAR, but I’m not sure why they would not be a rough predictor of success at uni.

    While the limitations of ATAR are well known, I have not seen a compelling case for replacing the current system, which for most school leaver admissions is based on ATAR but with room for taking other factors into account. It’s cheap to run, saves students the time and stress of a separate admissions test or process, and for higher ATARs seems to be a very good predictor of completion, which is more important than precise grades in most cases.

  7. She has a point…but in my particular case, I totally gamed the system.
    As a teenager more inclined to play sports, party, chase girls, etc, I wanted to get into uni, but with the least amount of work possible.
    So knowing that I wanted to do Economics and Commerce and that ANU is the best of the bunch, I reviewed their UAI scores before commencing years 11 and 12. A bit like the theory’s behind Michael Lewis’ Moneyball, ANU was totally undervalued in terms UAI’s. It was the best school in all the rankings, yet had the lowest UAI’s…probably because of its Canberra location…no not probably, I am sure of it !
    Anyway, I only needed about 75 to get in. I also knew that I only had to score this 75 from 3.5 subjects. So that’s all I did. I pi$$ed away certain subjects, specailised in certain subjects, and only did enough to pass the grade. I think I socred about an 80…although i think that score was downweighted as I went to a school of rough-niks which made our school not compare so well to others.
    Anyway, i got to uni, and immediately felt that my IQ was higher than an 80 score. But that said, it was true that i was a late bloomer. I spent my first 3 years at uni in the credit / pass grades, but then I scraped into the honours class, where I picked up with first class, some medal and a commendation….Smashed the aliens and those rote learning types right out of the park !
    I can’t say I remember much about that year, just that I learnt how to really learn. Yep, went in as a boy, and came out as a man. The rest they say is history.
    So yeah, late blooming does happen !

  8. low atar scores shouldnt matter. I got around 75% for all my subjects in the hsc even though I had a complication with my arm and was hopped up on morphine for the exam yet somehow u got an atar in the 30s
    whereas out of all the uni courses I was interested in the lowest entry choice was an atar of 66
    and in all honesty it doesnt seem fair that because I didnt know enough about history or english that I cant even get into the simplest of course and am now condemned to a low paying job

Leave a Comment