Guess the forever unknowable curve to the Feb 10 LSAT Forum

[tomwatts]

LSAC is usually ("usually" being the important qualifier here) pretty good about making sure that experimental sections are given far enough in advance or in different enough regions (international tests, etc.) that retakers don't see as a scored section a section which they previously saw as experimental.

I heard an LSAC test writer talk yesterday about the test, and he said that they don't like to call the experimental section "experimental," because "we're not experimenting on you!" Uh, then what else do you call it?

[Pricer]

Does that mean that there is a chance that people could have already seen these sections

It's pretty well guaranteed. The reason there's an experimental section on each student's LSAT is so LSAC can test (and curve) questions for future exams.

another question---many people have time issues with the lsat so if they go straight through (args for example) there is a good chanec that they will miss questions at the end, even if they are easy. any idea about how this is dealt with?

Learn which question types require the most time or are hardest. When taking the test, read the question stem first, and skip over the ones you know are hard.

[imisscollege]

Does that mean that there is a chance that people could have already seen these sections

It's pretty well guaranteed. The reason there's an experimental section on each student's LSAT is so LSAC can test (and curve) questions for future exams.

another question---many people have time issues with the lsat so if they go straight through (args for example) there is a good chanec that they will miss questions at the end, even if they are easy. any idea about how this is dealt with?

Learn which question types require the most time or are hardest. When taking the test, read the question stem first, and skip over the ones you know are hard.

No no I don't have trouble with this. I finish the sections -1 or -0 with time to spare. I just meant how does LSAC deal with this if they base the curve on people's question-to-question performance

[Pricer]

Oh. I figure that is just part of the test, and they probably just decided that it happens to everyone. The test is administered as four (well, six) sections, not as 101 (or 100 or ~126) separate questions.

[februaryftw]

No no I don't have trouble with this. I finish the sections -1 or -0 with time to spare. I just meant how does LSAC deal with this if they base the curve on people's question-to-question performance

From what I understand, they use pretty sophisticated methods for their curving. I think what they do is they analyze how test takers do on a particular question as compared to how they did on other questions. So, for instance, on question 24, even if they get 40% of students guessing, the distribution of who is getting the question right will still tell a lot. I also suspect that the numbers near the end of a section (in terms of how many people are expected to be guessing) are pretty regular from test to test, and if they aren't, you'll start finding that a lot of the students who are scoring lower on the earlier sections of the test are starting to get our hypothetical question #24 correct. This is just the start of an answer, but I think it is a big part of it.

[s0ph1e2007]

Ugh I dreamt it was -7 last night
One of the worst dreams ever lol.

[tomwatts]

They create the scoring grid based on performances on an entire section. If there's an easy question in the back of the section that most people miss because they don't get to it, it doesn't really matter; they just know that people who score a 175 on the rest of the test got 24 out of 25 on the section and people who score a 150 on the rest of the test got 13 out of 25 on the section and so on. They're not required to put the questions in order of difficulty, nor to know from question to question how a test-taker who has enough time to answer the question would do, just how well the section predicts your score.

As far as I know, they don't mix-and-match questions from different experimental sections, as, say, the SAT does, so they don't have to know how a particular question would perform out of context, just how it does in the context of the section that it's in.

[februaryftw]

For those of you who have yoru scores:

Do you have changed opinions regarding the curve?

[mazz]

this had to be a tight curve. like 7 or 6

[februaryftw]

I posted this in the mega-thread but I can guarantee this was not a -8 curve or anything close. I can almost guarantee 180 allowed for at least -2.

[trutherd]

I think it was pretty tight. After the test I tried to figure out how many I likely missed and where that put my score based on curves of -9, -11, and -14. I put my range at -12 to -16 and ended up with a 164, basically my worst-worst case scenario (164 is -16 on a -9 curve...)

[Pricer]

I'm thinking a tight curve of -9 or so. Maybe even -8.

[JasonR]

this had to be a tight curve. like 7 or 6

LOL

You serious, Clark?

[keg411]

Based on my score, I think us 2/20-ers had a -11.

[mazz]

this had to be a tight curve. like 7 or 6

LOL

You serious, Clark?

Did we sit for the same test? I didn't blow it, I hadn't gotten more than 2 wrong on RC in months, and was avg 1-3 wrong on LR for months. unless I got 20 wrong on LG (which i'm sure i didn't), I'm pretty sure the test was curved tightly, Lois.

[JasonR]

Wrong movie, but that's alright.

Has there ever been a -7 or -6 test? Anyway, every score from 170-180 except for 174 has been reported so far. Assuming that posters are being honest about their scores, that means -9 is the tightest it could be. More than likely, this test has the standard -10/-11 curve.

[existenz]

I think the curve was no tighter than -11.

The following scores have been reported:

180
179
178
177
176
175
173
172
171
170

IF a perfect score was required for a 180 AND nobody scored a 174, then that would be a -9 curve. But if somebody scored a 174 and you could miss one and still get a 180 (which almost always happens) then the curve is at least -11.

[JasonR]

Someone already reported a 178.

[existenz]

Someone already reported a 178.

Oops, mistyped. Fixed it.

[mazz]

yeah. -10 would be possible in that case. I must have totally blown the LG.

[dibs]

i really don't think it was that tight. i came out of the LGs thinking i was AT LEAST -10.

i have never done better than -3 on LR and RC. i was PT'ing around 153-155.

i scored 162. so either i completely rocked the shit out of LR and RC, or else everything i guess in LG was right.

*note* probably interesting to note that i only did my PTs while at work, constantly interupted by emails and phone calls, and never really sat down and did a whole timed test.

[robchuckbob]

I marked B for 10 LG questions and I still got a 166. I typically miss a couple in each other section, so I would imagine I missed around 16. So my guess would be that the curve was around -13.

[existenz]

i really don't think it was that tight. i came out of the LGs thinking i was AT LEAST -10.

i have never done better than -3 on LR and RC. i was PT'ing around 153-155.

i scored 162. so either i completely rocked the shit out of LR and RC, or else everything i guess in LG was right.

*note* probably interesting to note that i only did my PTs while at work, constantly interupted by emails and phone calls, and never really sat down and did a whole timed test.

If the curve was exactly the same -11 as June 2009, then a 162 could equal -22/-23 on the test.

In that case you could have:
LG -10
RC -4
LR -5
LR -4

And that's if you really DID miss 10 on LG. If you guessed on 10, you probably got 2-3 of them right, which means you missed even more RC/LR than my estimate above.

[existenz]

I marked B for 10 LG questions and I still got a 166. I typically miss a couple in each other section, so I would imagine I missed around 16. So my guess would be that the curve was around -13.

PT 57 and PT 56 had an -11 curve, and on those tests -16 equaled a 166. So not sure where you get the -13 formulation from.

All of this is just wild speculation by all of us, including me. None of us, not even those who got 180, really know how many we missed.