Why didn't my MGMAT test adapt more?

[NGL]

Yesterday, I scored 45 on the quant portion of an MGMAT practice test.

I mismanaged my time and had to rush on questions 30-37. In turn, I got every one of those questions wrong. However, questions 13-36 were all 700-800 level. Question 37 finally dropped down to 600-700.

Given I had a string of 8 questions wrong, why didn't the test adapt down more/sooner?

[StaceyKoprince]

The test was still adapting as you went. :)

Multiple possibilities. First, the database is allowed to choose from a broader range of levels than you might think because it also has to balance various other parameters - a certain # of PS vs DS, a certain number of geometry overall, a certain number of triangle vs. circle Qs, etc. It's not only about the difficulty level.

Second, the data that you're seeing in the report is not necessarily 100% accurate / what's in the database. When we first post a question, we assign a difficulty level to our best-guess estimate. As people do the questions, we track the data and assign a real difficulty level based on actual results. The database difficulty values then get updated, but the field that shows you the difficulty range has to be manually changed in a completely different way and by a different team than is handling the assignment of the real difficulty values - so there is always lag there in updating the "viewed" difficulty levels vs. the real, new levels assigned. You could have had a few of those.

Finally, the database is not allowed to choose questions that you've seen before (in the same set of 6 tests), so that limits the options as you take more tests, particularly if your scores aren't changing that much. (I'm guessing this isn't what happened in your case, though, because usually the consequence here is that someone "uses up" the 700+ question pool by scoring very well from the very first test, and so they drop to 600-700 sooner than expected - but that's not what happened to you. I'm just including this possibility for others reading this.)

Now, let's talk about the more important issue: what are you doing about that timing problem?

Read these and start doing what they say! :)
http://www.manhattangmat.com/blog/index ... to-win-it/
http://www.manhattangmat.com/blog/index ... nt-part-1/

[NGL]

Thanks, Stacey. I've been trying to improve my timing for awhile now, but somehow I really mismanaged it during this test.

I think one of my biggest issues is that I haven't developed a "1-minute sense." I've been using benchmarks, and I am comfortable with guessing on questions that are too hard for me, but there are always around 4-5 questions that I unintentionally end up spending 3+ minutes on.

On this most recent test, I spent 3+ minutes on 8 questions; I got 5 of them correct. So, I need to figure out out to solve them more efficiently.

[StaceyKoprince]

Good - that's your next step. First, make sure you got them right because you really did know what you were doing (there might have been one or two that ended up being educated guesses or even pure luck). Next, figure out whether you can actually get the time down to 2.5 minutes or faster - and, if so, how. What do you need to do / study / learn / practice in order to make that happen next time around?

Note: for some, you may decide: hmm, getting this right in 3.5 minutes is not worth it - I'd rather get it wrong in 1.5m (pull the plug), make a guess and move on, and spend that saved 2 minutes elsewhere.

In particular, if you ran out of time, rushed, had to guess blindly at the end, or made careless mistakes anywhere along the way that could be attributed to speed - then that 2 minutes is better spent in those other places. :)