Is the 7Sage simulated Flex PT easier than actual Flex?

An_Skywalker

Ever since the October test became Flex, I've been PTing using the "simulate flex" option on 7sage. However, I'm wondering if this will give me an inaccurate picture of how well I'm actually doing, since I don't know how 7sage decides which of the 2 original LR sections to pick, it could well be that I've been doing easier LR sections than what will actually appear on the test.
This being said, is there a good estimate out there of how the 7Sage simulated Flex scores will correlate with the actual Flex? Many thanks!

cantmakeupmymind

7Sage Flex is just picking the first of the two LR sections - so if the LR sections were 2 and 3, you'll take the section 2 LR.

For the Flex, LSAC is using existing LR sections from non-disclosed tests. As far as anyone can tell, they aren't mixing and matching LR questions from different sections to create a Frankenstein version. On the test day, you may get an "easier" section or you might get a "harder" one - it just depends on what LR section you get of the handful they've decided to use, which is partially based on the time you schedule your test. They're swapping different sections in throughout the weekend/week as people are testing - there's not one August Flex, one October Flex, etc. There are lots of different versions, so the test that someone takes at 10am on October 3rd could be completely different than the one someone else takes at 4pm on October 4th (or it could have sections in common).

cpeaks13

So if you are taking the flex should you definitely be using the simulate flex method or just PT'ing normally? #help

Logician

@cpeaks13 I would just carry on doing full 4 section PT's. You can always calculate your flex score after based on either LR section (I'd calculate it based on the lower one). Not to mention this will also be great for your stamina come actual test day.

hopefulling

I'm wondering if this will give me an inaccurate picture of how well I'm actually doing

Yes, for this reason! I don't take the PTs as Flex because of the range. I miss a few more questions usually on the 2nd LR compared to the 1st and the range would have been such an incorrect assessment of my true preparedness. The range is sometimes 4 points and if I were to receive that lower score on an actual test after seeing the higher score, I would be devastated. Plus, it would have pushed me into a higher Phase of preptesting before I was actually ready with my average timed score.

Plus, like stated above. It feels nice to be over-prepared for a 3-section test than leave those learning opportunity questions on the table (there's plenty of drill questions in tests 1-35). Right now, I blind review every question and it might add something like an extra hour-and-a-half into a test-day.

I've also noticed that some LR will have the first section heavier in a few question types and then use the 2nd LR for other question types. It might also under-estimate or over-estimate performance in some question types.

calvinnhobbes

I'm sticking with full PT's for the stamina reason. I feel like the more questions I get exposed to before test day, the better. I did take one Flex-simulated PT just to see how it felt time-wise and mentally which was helpful to know though so I would suggest doing that if you want to feel out the format. Good luck!

An_Skywalker

@summerofstudying said:
7Sage Flex is just picking the first of the two LR sections - so if the LR sections were 2 and 3, you'll take the section 2 LR.

For the Flex, LSAC is using existing LR sections from non-disclosed tests. As far as anyone can tell, they aren't mixing and matching LR questions from different sections to create a Frankenstein version. On the test day, you may get an "easier" section or you might get a "harder" one - it just depends on what LR section you get of the handful they've decided to use, which is partially based on the time you schedule your test. They're swapping different sections in throughout the weekend/week as people are testing - there's not one August Flex, one October Flex, etc. There are lots of different versions, so the test that someone takes at 10am on October 3rd could be completely different than the one someone else takes at 4pm on October 4th (or it could have sections in common).

Thanks! I was really confused about how 7Sage picks the LRs.

nikkiwantsa160

About to take my first practice test with the Flex and I'm nervous! I know it's only the first but it can be so discouraging if it doesn't turn out well

cpeaks13

@nikkiwantsa160 would love to hear how it goes!

cpeaks13

Thank you @Logician that is what I was thinking as well!

An_Skywalker

@nikkiwantsa160 said:
About to take my first practice test with the Flex and I'm nervous! I know it's only the first but it can be so discouraging if it doesn't turn out well

you'll be fine! focus on perfecting the LG as much as you can! LG is the easiest section to improve

lilpingling

I've taken several PTs in which I first simulated the flex and scored, then immediately took the additional LR section and added in the results and rescored. The most my score has ever deviated was a single point. The question types vary, but it seems the skills being assessed remain fairly consistent. I've done everything I can to create a disparity and I can't make it happen. This may be an unpopular opinion, but I think it's pretty similar in the end.

VerdantZephyr

@lilpingling I think that is atypical. I first started calculating my possible flex scores while doing 4 sections. I went something like -0 LR, -1 LG, -4 RC, -5 LR. Those are very different Flex scores depending on which LR section is used, checking against a couple of random numbered tests about a five point swing. That is a major deal. I do not know if this is true or not, but it may even be that older tests had a harder and easier LR section, but that the two sections together had an average difficulty equivalent to other LSATs.

lilpingling

@VerdantZephyr said:
@lilpingling I think that is atypical. I first started calculating my possible flex scores while doing 4 sections. I went something like -0 LR, -1 LG, -4 RC, -5 LR. Those are very different Flex scores depending on which LR section is used, checking against a couple of random numbered tests about a five point swing. That is a major deal. I do not know if this is true or not, but it may even be that older tests had a harder and easier LR section, but that the two sections together had an average difficulty equivalent to other LSATs.

Wow. That's really significant. You've definitely convinced me to continue using both sections moving forward.

ninamatryoshka

@VerdantZephyr said:
@lilpingling I think that is atypical. I first started calculating my possible flex scores while doing 4 sections. I went something like -0 LR, -1 LG, -4 RC, -5 LR. Those are very different Flex scores depending on which LR section is used, checking against a couple of random numbered tests about a five point swing. That is a major deal. I do not know if this is true or not, but it may even be that older tests had a harder and easier LR section, but that the two sections together had an average difficulty equivalent to other LSATs.

maybe this is because you're tired out by the time you get to that last LR section and so you're missing more questions, and not because the section itself is more difficult?

J.Y. Ping

Hi everyone, this is a great thread and I just want to point your attention to @"Alan Cheuk"'s comment from this thread that addressed the core question of how we chose which LR section for Flex and why.

We ran numbers on a large sample of actual student PrepTests. We pretended they did Flex by (A) skipping first LR section or (B) skipping second LR section, applied our Flex Score Convertor, and looked at the average difference in scores.

We found that the spread between the averages for A and B was small. Across the 95 PTs, the biggest spread was a 1.38 (PT 9), and 78 of the 95 PTs had spreads of less than 0.7. For reference, the standard error of the actual LSAT test is 2.6 (https://www.lsac.org/lsat/taking-lsat/lsat-scoring/lsat-score-bands).

That is why we chose the uniform rule of removing the second LR from each PT for Flex simulation. It keeps things simple, and the difference between A and B are small relative to the standard error of the test itself.

To be more explicit, Alan is saying that yes, there is a difference depending on which LR section you take. But that difference is less than 1 point.

@lilpingling Your experience is statistically typical.

Also in that same thread:

According to what LSAC has told us, if you did a standard 4-section test, that is the best predictor of your Flex score. Better than any Flex calculator, ours or otherwise.

@calvinnhobbes @Logician

kadenktnk

JY saying that the difference is statistically insignificant but then saying that the standard 4-section is the best predictor....now im really confused

willyuu_

I have actually been practicing exclusively with the Flex format since August and it has helped me understand how crucial doing well LG and RC is in Flex compared to the standard test. RC or LG questions basically worth more now and it might these sections more stressful.

For that reason alone I would recommend doing a few Flex’s.

VerdantZephyr

@kadenktnk said:
JY saying that the difference is statistically insignificant but then saying that the standard 4-section is the best predictor....now im really confused

Because taking two sections means that your performance is averaged. Even if one section happens to be an outlier you'll receive the average of your two scores which, given that any two given LR sections usually have a difference of 1 or less, is going to be closer to your typical score. Outliers are not a good way to track progress and a flex score, even if it is similar to your test day experience, is more prone to outliers. It is the same idea, though on a smaller scale, as scoring 163, 164, 163, 157, and 162 not panicking over the 157. Same with getting too happy about a 170 when your average is 161. Taking the four section and comparing the scores is also like taking an extra flex in some ways because you can generate two flex scores from it.

Jay Tee

I took my first flex PT and scored the same as I did on my previous real (5-section) take for whatever it’s worth — I think the scoring will be consistent as JY says. 🤷 Try alternating between both versions and see what happens.

Alan Cheuk-2

@VerdantZephyr said:

@kadenktnk said:
JY saying that the difference is statistically insignificant but then saying that the standard 4-section is the best predictor....now im really confused

Because taking two sections means that your performance is averaged. Even if one section happens to be an outlier you'll receive the average of your two scores which, given that any two given LR sections usually have a difference of 1 or less, is going to be closer to your typical score. Outliers are not a good way to track progress and a flex score, even if it is similar to your test day experience, is more prone to outliers. It is the same idea, though on a smaller scale, as scoring 163, 164, 163, 157, and 162 not panicking over the 157. Same with getting too happy about a 170 when your average is 161. Taking the four section and comparing the scores is also like taking an extra flex in some ways because you can generate two flex scores from it.

VerdantZephyr is spot on. We saw little variation in the difficulty between the two LR sections, so any student's expected score is similar for both sections. However, there is significant variability in a particular instance of testing. Averaged over multiple prep tests, there is little difference in the average, but there is more variance in the scores. 4 section tests reduce the variance. Another way of saying this is that 3 section flex has good accuracy but poor precision.

estradadaniela55

Does anyone know how LSAT flex works? Do you do each section back to back. Or do they give you some minutes in between each section?

An_Skywalker

@estradadaniela55 they give you one minute in between each section, and the next section automatically starts once the clock runs out.

estradadaniela55

@ashen03 THANK YOU for responding. I can add that to my prep!