Did UMAT scaling change this year...?

Hey everyone... sorry if this sounds really dumb BUT some friends and I were having a look at our UMAT results from 2011 and 2010. One of my friends got the same raw mark as me in '10 (with 59) but he got 89th percentile... I only got 83.

What gives? Does this mean that the raw scores required for interviews are also going to skyrocket? By the UNSW (first preference) site I needed a raw score of 59 to secure an interview with my ATAR of 99 but because of percentiles there'll be another 6%-better-UMAT's worth of people before me now won't there?

Its because percentiles relate to how well everyone does that year. So its natural the raw scores will not always correlate to the percentiles.

I'm taking it no university would even bother to look at a person's percentile then seeing as they can correlate to completely different raw scores between years?

Actually a lot of universities consider percentiles rather than raw scores. The percentile is more accurate giving how well a candidate did in a year. They dont correlate it between year, and thats another reason why the UMAT is only used for two years.

Originally Posted by miss_universe

They dont correlate it between year, and thats another reason why the UMAT is only used for two years.

It surprises me that they don't scale the raw marks to match percentiles each year, I mean, one would think the sample quality would not change significantly year to year. That said, it's the black-box that is ACER so who knows what the real reason is....

But how would you standardise the difficulty of the test each year?

Originally Posted by Jackie

But how would you standardise the difficulty of the test each year?

Statistically speaking, IMHO they wouldn't need to, it's a large enough sample size. I guess I'm just surprised they release supposed 'raw scores' when it would make more sense just to release 4 percentiles.

I know one can work them out roughly from the graphs, but it would just seem more logical if they're not going to scale similarly each year. Ultimately, irrespective of the method used by the uni... it's one's rank that matters, and should be year vs year i'd think too. Be interesting to know how the unis that deal in 'marks' as opposed to percentiles do it.