siddharth_
@bb
DI Mocks data
1.
https://gmatclub.com/gmat-focus-tests/r ... 27685.html -- new, 74, 13 correct
2. https://gmatclub.com/gmat-focus-tests/results-2322982.html -- old, 79, 10 correct
3. https://gmatclub.com/gmat-focus-tests/results-2322641.html -- old, 82, 12 correct
4. https://gmatclub.com/gmat-focus-tests/results-2322047.html -- old, 85, 16 correct
2 questions --
1. the new algo mimics the scoring logic more closely or also the question placement logic?
2. the algo is throwing similar questions all at once -- can this happen in the real test too? (All DS questions bundled at the beginning, no MSR till the 16th question, even in Verbal 4-5 BF at back to back)
Hi. In general the new algorithm is supposed to be more accurate - that’s the big picture. Today thanks to another user in this group,
Rishm, I learned that a small number of new HARD questions were not marked as experimental so you got 1 of them served as well while being considered a medium level (this is addressed), so this had probably 1-2 points impact on your test.
In general, looking at your test, the old algorithm, you had a lot more hard questions (5) vs. this one had only 1. In the latest test you got 2 easy questions and you got one of them incorrect. You were doing well until you ran out of time. But I think the approach of guessing the last 5-6 questions is not the best approach. Here is a chart of your score:
siddharth_
the intent of asking these questions is - while these tests are good for practicing but how closer are the scores to the real thing?
I was thinking of booking the test this week, but with this mock result, the confidence level came down a little.
I would Highly suggest taking a GMAT Prep to establish the test accuracy - GMAT Prep is super helpful for doing that and being a second data point to confirm/reject. You should have gotten a higher score but one question torpedoed you. We should have some updates coming this next week (including spreading out the questions by type - right now the system picks the best questions and there are a lot more DI questions in the database they tend to be the closest ones to pick at first, and then it starts picking the other questions that match the parameters). It is working correctly technically speaking but that results in some cases into a stacked list of questions where it serves each question by type - something we learned only earlier this week as this also did not come up in testing... so I would suggest waiting another week to book your test. Ideally take a GMAT Prep and if you don’t have a spare GMAT Prep test, wait until Thursday or Friday to take another test as by then the order of questions will be more natural (though it will be a bit forced but that’s the more appropriate approach. Just in case, I have extended your subscription by a week just in case. But meanwhile, use this result to motivate yourself to study harder and be a bit more paranoid - that will only help!