1

**General Off-Topic Board / Re: Feb vs July bar exam scaled scores**

« **on:**August 01, 2014, 03:40:06 PM »

I agree that there is a lower pass rate and thus "harder" on paper but I guess what I'm asking is this.

Since the average taker is a retaker (who statistically do worse) and/or a foreign trained attorney (same type of stats) wouldn't that lower the curve for the otherwise average person?

For example, if I showed up to a class in Algebra that was filled with remedial math students

that was on the curve I'd get an A even though I am most likely not an A student in math otherwise.

That is the angle I was looking at it from.

You'd be exactly right if the scale worked like a curve but it's slightly different. The way the MBE scale works is that it takes a subset of about 20-30 questions on each exam iteration and studies how students did on that subset as compared to how they did overall. The subset is comprised of easy, medium and difficult level questions and these same questions have been repeated (with minor variations) from year to year since the MBE was first administered. Students who performed poorly on the MBE should, in theory, perform poorly on the subset of questions. Students who performed so-so on the MBE should, in theory, perform so-so on the subset and likewise students who aced the MBE will ace the subset.

If enough students who aced the MBE have trouble doing equally as well on the subset, then the exam is "scaled" up to compensate for the difficulty of the exam. Similarly, if enough students who bombed the MBE unexpectedly end up acing a significant amount of the subset questions then the exam will actually be "scaled"down. The theory is that by using this scaling process every MBE that is given each year should not be any harder or easier than the first MBE that was ever given way back when.

So even though February does have a large % of retakers, and retakers do tend to perform poorly, the test will not be scaled up or down just because the retakers perform poorly. The scale already accounts for that. In order for the scale to move the needle one way or the other, a lot of smart kids have to do well on the MBE as a whole while simultaneously screwing up on the subset (meaning that the exam washarderthan previous years which will result in a bumpupfor everyone), or alternatively, a lot of not-so-smart kids have to screw up on the MBE as a whole while simultaneously acing the subset questions (meaning that the exam washarderthan previous years which will actually result in a bumpdownfor everyone).

I know, it's crazy right?

Hey Burning Sands!