How do you know if a test is harder/easier than usual? Forum

[DorianGray89]

I've been scoring in the low 160s (160, 162, 164) and yesterday took PT 56 and got a 171. Not sure if it's because it's an easier test or just because I'm improving.

So, how do you know if a test is harder or easier than usual? the conversion table?

[Eugenie Danglars]

If you're trying to improve, track your raw score (the # right) instead of the scaled since, as you noticed, the scaled can vary a lot.

To tell if a test is easier or harder, look at how many you can miss and still get a 170. If you can only miss 8 (ie, a 92/93 is a 170) it's harder than if you can miss 13 (ie, an 87 is a 170). These numbers are just hypothetical, but you get the point.

[gendefect]

But isn't the whole idea behind the scale that the scale exists based on the difficulty of the test?

In other words, if you can get a 170 with 8 wrong on one test and a 170 with 13 wrong on another, the latter test isn't easier because the scale lets you get more wrong, it's actually harder, in terms of the difficulty of the questions that are asked, and the scale is adjusted based on that difficulty. That would mean that, theoretically, because of the scale, all the tests are of equal difficulty.

That said, it seems to me that the tests have gotten more difficult the more modern they get, but maybe that's just me, and either way, it would be a gradual progression. And even if that is the case, it would speak to the OP as having actually improved as evidenced by PT 56.

[St.Remy]

But isn't the whole idea behind the scale that the scale exists based on the difficulty of the test?

In other words, if you can get a 170 with 8 wrong on one test and a 170 with 13 wrong on another, the latter test isn't easier because the scale lets you get more wrong, it's actually harder, in terms of the difficulty of the questions that are asked, and the scale is adjusted based on that difficulty. That would mean that, theoretically, because of the scale, all the tests are of equal difficulty.

That said, it seems to me that the tests have gotten more difficult the more modern they get, but maybe that's just me, and either way, it would be a gradual progression. And even if that is the case, it would speak to the OP as having actually improved as evidenced by PT 56.

+1 to all of this except for the last line. The 171 sounds like it is probably a fluke thing, you get some outliers taking LSAT PTs. To OP, track your score since that's what you should actually care about, and in terms of telling what test is "easier" or "harder" it's going to be a personal thing, so go with your gut. If you run across a test where all of the games click and very few of a question type that you find difficult show up then that will probably be an easy test for you even if it probably wouldn't be for some other people. With the LSAT there is no single number, raw or scaled, that will tell you how difficult the test was for you.

[gdane]

Im confused... I thought the looser the curve, the easier the test and the tighter the curve, the harder the test. If a -13 gets you a 170 on one test then this test was presumably easier.

I look at my performances as confirmation of this theory. Two days ago I scored a raw 79 and I got a 161 and yesterday I scored a raw 75 and got a 161. Im not sure that yesterdays test felt "harder", but I can definitely say that the test before definitely felt easier.

Whatever. For me the goal is get as many correct. F* the curve!

[094320]

..

[fosterp]

They scale the score so that scores can be comparable across different tests. Since its impossible to exactly match the difficulty, you can't say that a 170 on one test is "exactly" the same as 170 on another. So each test has its own scale. A hypothetical situation, someone that is at a true 170 takes test A, and the test is a bit harder than average, and ends up getting a lower raw score he/she usually does and gets, for example 87. According to the average curve on the more recent tests, that would place him, on average, around a 168. However since this test was determined to be slightly harder, lsac will adjust the conversion chart and place this test taker at 170. Likewise, the same person takes test B and the test is a bit easier than average, and he ends up getting a higher raw score than average, say 93. Likewise the conversion will reflect the difficulty of the test and that much higher score will still convert to a 170.

Of course, difficulty is relative and we all have our own strengths and weaknesses, but I believe they can and do assess average difficulty as questions are completed as experimental sections. In general, you can assume that a "tighter" curve (meaning you need a higher raw score for a given lsat score) would be easier than a test with a looser curve. Also I believe they do assess tests not just in their overall difficulty, but also in the difficulty for a certain pool of test takers (difficulty of the test to people at the 160 level, and difficulty of the test at 170), so test A might be harder than test B for 160 scorers, but test A might be easier than test B for 170 scorers. The curve will reflect that as well.

Given that, you can say that getting a higher score means you simply did better. The test difficulty should be reflected in the score conversion.

http://lsatblog.blogspot.com/2010/02/ls ... -lsac.html gives you a good thorough explanation

[whymeohgodno]

If it has snakes+lizards, it's the hardest test.