How ACER calculates UMAT scores

Hey Guys... was wondering if there is any info on how they calculate UMAT raw scores... I'm assuming standardization looks at how the overall population did on each question based on which a weighting is applied to each one? Meaning you can't work out how you are likely to do based on the prac booklets from Acer?

I HATE section three... only got about 2/3 of it right in the prac booklets.

Also... anyone who is looking to increase their section one score... try a speed reading course. A mate of mine did this one he bought on Amazon... trains you to read faster and retain more!! Best part is you can load your own material into it (so anything technical you find on the internet...)

EDIT: I haven't used this system, so check out the functionality properly and don't rely on what I have put here... cheers!

"Speed your Read"... it's not a UMAT specific product so I should be able to post this here right?

http://www.amazon.com/Speed-Your-Read-Read...83431704&sr=1-1

It's all a BIG secret. There's a lot of speculation that others might be able to fill you in on but ACER doesn't like to tell anyone how they calculate the raw scores. It's fairly statistical though, lots of analysis of questions etc.

Yeah, section 3 was hard for me too, give me 1 and 2 anyday. I've got a theory that girls are better but there's plenty of exceptions to the rule.

Speed reading is popular among UMAT candidates... I don't know, though

Thanks!

I am all for speculation!! Gimme some... or please direct me to the post that has it already (couldn't find it in here by searching)

Originally posted by animadverted@Jul 3 2007, 01:32 PM
Yeah, section 3 was hard for me too, give me 1 and 2 anyday. I've got a theory that girls are better but there's plenty of exceptions to the rule.

[snapback]47360[/snapback]

LOL way to prevent a battle of the sexes debate there :P

OK, well the UMAT scores are scored in such a way that they should be equivalent from year to year (i.e. a 74 in section 1 in 2004 = 74 in section 1 in 2006). ACER says they don't weight questions differently but I don't believe them lol. There's a huge variation of UMAT scores within percentiles (i.e. a 99 percentile can be a UMAT score in 60s, 70s and 80s). I don't think UMAT scores are equivalent across sections (i.e. a 74 in section 1 isn't the same as a 74 in section 3) but I have NO idea if that's true or not. Ummm... yeah basically we don't know lol.

I believe for UMAT 2005, there was a document on MSO which tabulated people's percentages and subsequent scaled scores. No one made a document for 2006. It could give you a BRIEF idea of how percentiles convert to scores.

They are generally consistent. in UMAT 2005 i got section 3 - 100%tile and was scaled score 80. UMAT 2006 was again 100%tile and scaled score 81. (I only know of people getting 81 or 94 i believe).

94 score in section 3? dayum

i got:
section 1: 99percentile, 74 score
section 2: 86percentile, 57 score
section 3: 99percentile, 73 score

it's sad i can still remember that.

I'm sure this has already been discussed somewhere in the forum, but I'm teh n00best. Cut me some slack.

I'm applying for B Med at the University of Newcastle for 2008, and am registered to sit UMAT in July. I read that a threshold score of approximately 180 is required. I just wanted to clarify that this 180 is a score out of the total 300 (for each section, totalling 100 each). At first I was a bit surprised - after all, 180 would only be about 60%.

Is there more involved? Like, I know that UMAT keeps its scoring strategy guarded in Top Security Vaults, but whatever.

thanks

Hey,
Let me put it this way: My percentile for section 1 was 99 (i.e. I was in the top 1% of all UMAT applicants for that section) and my UMAT score for that section was 74. So even being in the top 1% for a section, 74 is a really big score. It's not the same as say 74% in a normal test.
So yes, each section is out of 100, but it's actually quite a big feat to score say 60 in each section.
UMAT scores are standardised to be equivalent across years. I'm thinking there must be a mean and standard deviation thing going on but IDK.
Last year the cutoff for Newcastle was 187 for "normal" applicants. If you live in or around Newcastle and/or are rural, then the cutoff is lower.
Once you've got an interview you've got a 1/8 chance of having a non-bonded CSP place.
Also, check the Newc website, as I think there's form you need to send directly to the faculty apart from your UAC application.
Good luck

Why sad? I can remember the percentiles and if I tried, I probably could get the scores too.

The people that spit out crap like "If you can remember your enter after one year at uni, you're pathetic" irritate me like nothing else. It's a four digit number... and oath, I worked hard for it, of course I'll remember it

/rant

Further proof that we QLD'ers are slackers, we only have to remember a one digit number - yay for the OP system

Well, I guess if you get like an OP10-25, you'd have to remember 2 digits, but if you ended up getting such an OP, I think that'd be beyond your capabilities anyway :P