Page 1 of 1
Are older tests even a good indicator because of the curve?
Posted: Sat Mar 15, 2014 6:45 pm
by Darmody
delete
Re: Are older tests even a good indicator because of the curve?
Posted: Sat Mar 15, 2014 6:50 pm
by quigglyboom
Judge it on the score, even within a single year there is a marked variation between the number of correct answers that correspond to each respective score.
Re: Are older tests even a good indicator because of the curve?
Posted: Sat Mar 15, 2014 6:52 pm
by Nova
go by the equated score.
the curve for 170 has fluctuated from like -8 to -14 in recent years based on the difficulty of the test.
Re: Are older tests even a good indicator because of the curve?
Posted: Sat Mar 15, 2014 7:36 pm
by Jeffort
To get the full benefits from taking practice tests you shouldn't put too much weight into either the scaled or raw score, but instead just look at it as a rough gauge of current score range.
Instead of celebrating occasional outlier scores or sudden jumps, patting yourself on the back and risking getting overly confident/complacent in your current abilities, you should instead honestly evaluate how many of the points you earned legitimately from questions you got right because you fully understood the logic and CR vs. questions you got lucky on but really didn't know or understand why the CR was correct when you selected it instead of a trap for whatever lucky reasons of the moment.
If you count in the points earned from lucky breaks that went your way on questions you had down to two but weren't positive which was correct when you made your 'educated guess', you'll end up with an inflated belief of what your actual ability level really is. For each one of those where your 'gut instinct' choice was correct, there are others where it wasn't. There is no guarantee the lucky breaks when down to 2 will go your way more often than not on test day like it does on some PTs when under much less mental pressure and stress.
It's much more important to look at PT scores critically to flush out weaknesses and identify all the questions you got right partially by luck or instinct rather than full understanding since more of those questions you are weak on tend to go the wrong way on test day due to test day environment, extra stress and pressure.
It's important during review to look at all questions, not just the ones you got wrong, to check and see if you got questions correct for the right reasons. Once people start taking the time to carefully review questions they got correct on PTs in addition to wrong ones, they quickly realize that they've been getting a lot of questions correct for completely wrong reasons and really don't understand the logic as well as they thought they did. Better to figure this type of stuff out with deep honest review well in advance of test day instead of just relying on predictions made from highest PT scores that may have partly been the result of some luck and a few more instinctive gut reactions/decisions breaking in your favor than usual on hard questions.
Re: Are older tests even a good indicator because of the curve?
Posted: Sat Mar 15, 2014 7:39 pm
by northwood
like other posters said- go by the actual scores, not the scaled scores to gauge your improvement. Aim to get the most correct responses in each section- and categorize the ones you keep getting wrong ( both location in the section and question type) and if a trend develops ( there usually will be one) go back, review, and reassess those areas ( but don't neglect the other question types)