***Class of 2016 NMSF/NMF Qualifying Scores

Paying for College National Merit Aid
361

@abbylawi @rhubarbb … Just to clarify a few things, the Commended score is a national cutoff (same for everyone regardless of state) while the Semifinalist scores vary from state to state. None of these scores will be known officially until September when NMSC has their public release date, but info starts to leak out earlier than that as letters are sent out to school principals.

What happens here on CC is that some homeschoolers (whose parents are their “principals” and thus get access to the letters earlier than others) and a few cooperative school principals start posting as they receive their letters. For the Commended cutoff, letters were sent this spring to schools to “confirm the status” of the top 50,000 scorers from the PSAT and that letter supposedly listed the Commended cutoff as 202. This info has been very accurate in the past, so I would feel pretty confident that 202 will bear out as the national Commended cutoff.

As late August/early September arrives and homeschool parents/principals start receiving Semifinalist notification letters, CC members start “bracketing” each states cutoff scores. These letters don’t specifically mention the state cutoff score, but posters here start the process of “Student A is a Semifinalist with a score of XXX or Student B is not with a score of YYY” which starts narrowing down the range which that state’s cutoff falls into. Historically, CC posters tend to have the vast majority of states figured out correctly before NMSC’s national release date (usually around September 10th-ish).

If a school counselor is telling you that they know a state cutoff score already they are mistaken. There are several threads here on CC with a wealth of information on the NM process, and folks here are always willing to help as well.

362

@rhubarbb that was likely the cut off for your state last year.

363

We live in NY. My daughter was mentioned as a Nat Merit soemthing, but she was taking her AP so she isn’t sure, but probsbly just commended. Her score was 209. Her friend who got 211 also made the cutoff.

364

I believe that someone had called NMSC in last month or so and that they said the commended cutoff was 202.

365

Can anyone predict what Maryland’s cutoff will be for NMSF this year?

366

@Chalmydia

See post #293. Unchanged from last year 221 is the most likely outcome. The goodness of fit for MD wasn’t high on the regression analysis, however.

367

Is 202 higher than usual for commended? I thought it is around 200 each year. Anyone want to guess if 223 is safe
for VA?

368

@longdog

Commended has ranged from 200 to 203 from 2005 to now. Last year (2013 test takers) it was 201. The year prior it was 203. So 202 is on the higher side, but down from the highest. In general when I have calculated my estimates for this year for the various states NMSF cutoff, they are, in general, similar to last year or +/- 1, especially for the big states.

If you need or want something more precise for VA, I am willing to go through the calculation for VA if you do some legwork: I need the number of students for VA who scored in the 75-80 range for each of the 3 subject areas. I also need the total number of test takers. I will need that for each of the years 2008 through 2014. To get this information click on http://research.collegeboard.org/programs/psat/data/cb-jr/archived/2013 and select the proper year then click on the state report for VA. It will take about 5 or 10 minutes.

See post 348 for a nice example compilation of data from another state.

369

@WRUAustin
I would really appreciate an estimate for Louisiana.

2014-15 / 99 / 130 / 126 / 11281
2013-14 / 77 / 125 / 89 / 11703
2012-13 / 78 / 98 / 104 / 11701
2011-12 / 71 / 77 / 145 / 12453
2010-11 / 83 / 110 / 134 / 12104
2009-10 / 83 / 113 / 81 / 12622
2008-09 /86 / 66 / 89 / 12340

Thank you.

370

@TheTherapist

The correlations are generally better with states with larger populations and also states with higher cutoff scores. Louisiana doesn’t have either one. Also, the spread in cutoff scores was only 3 and the one 210 point (the high one) didn’t fit the curve well. So, the R2 (goodness of fit) wasn’t high. Still, I think the calculations may be helpful. For the previous year, the calculations would have predicted 208.8 and the actual was 208. For this year the prediction is 209.9. So, my best guess is 210. Next best guess is 209. I hope that helps and is what you want to hear.

371

@WRUAustin
210 is NMSF hopefully no higer. Thank you again and have a great Memorial day weekend everyone.

372

Despite all the calculations that I have done that show my S with a 1 point margin in Texas (219 score with likely 218 cutoff), I sometimes have doubts. My calculations use the individual state reports and I use the scores from 75-80 on all 3 sections combined with the growth in state populations and the past cutoff scores to perform my linear regression. (In Texas I also used the actual public school enrollment numbers which show a relatively big jump in enrollment for this years junior class over the last year and the R2 improves using this data instead of general state growth).

One thing that was troubling me a bit was the 1 point increase (to 202) in the commended score over last year. The biggest difference between commended and cutoff from Texas has been 17. If that increased by 1 then the cutoff would be 220 and my son would be out. So, I decided to look to see if there was a correlation between the numbers who scored from 70-74 compared to 75-80 between years. There is no reason to believe that the distribution of scores is the same from year to year.

So, I summed the 3 components numbers of those who scored 70-74 and divided them by the numbers that scored 75-80. What I found is that when that ratio is low (less than 1.85) the delta between commended and TX cutoff was high (17). When the ratio was high (greater than 2), the delta is more likely to be 16 or 15. This makes sense. If the number of good but not great scores increases, the commended score will be higher. But these good but not great test takers aren’t making the state cutoff. And the R2 values for relatively higher scoring states are consistently better with only the 75-80 scores being used. For this year in TX, the ratio is high and points to a difference of 16 and not 17 and certainly not a new high (in delta) of 18.

The TX cutoff scores had trended higher so the more recent data is more useful than the older data.
Test taking year,ratio,delta
2014,2.09,?
2013,1.86,17
2012,1.96,16
2011,2.15,16
2010,1.31,17
2009,2.29,14
2008,2.29,15

So, for relatively high scoring, populous states, the fact that the commended score rose by 1 point does not necessarily indicate that the cutoff score is likely to rise for your state (assuming TX is a good guide and the TX data is generally well behaved from my analysis). For low scoring states a rise in the commended score could be more noteworthy.

So, after today’s data obsession, my best TX guess is still 218 and I feel confident that 219 is relatively safe.

Now I ideally will wait until early September…

373

My DS is sitting on a 220 in Texas. I feel fairly confident. Thanks so much for all of your analysis.

374

Does anyone know the score required on the SAT to keep NMF status?

375

@hsmomof3

See post #358. There is a more clear explanation elsewhere, but this one is mostly clear.

376

@WRUAustin

I like looking at your analysis and your thoughts for the 70-74 bracket make sense. I would assume that for lower scoring states putting both brackets into a correlation would help the R2. The increase in the commended definitely was predictable when comparing both brackets this year v last. I think 219 is awfully safe…although I feel more confident with my daughter’s 218 in Indiana. Based on looking at several data sets, it looks to me like the Math score often has less of a correlation than the other two…maybe there are a lot of math wizzes that aren’t so good on the other sections? Maybe you have already thought about the possibility of getting a better fit if the weighting of the math component is lowered slightly. Thanks for your work.

377

@chargerparent

Thanks for your thoughts. Nice to have someone speak geek to me. I had thought of putting a weighting on the math, but I had never done it (I was one of those kids who scored high in math, but not the others). Your post spurred me to look at it. Thanks. I figured I would give it one constant weighting. Changing it year to year wouldn’t make sense. So, I gave the math a 0.6 weighting based on what I saw over the 7 years of data I am using. After doing this the R2 went from 0.86 to 0.93 (which is pretty darn good, I think). The TX prediction went from 217.99 to 217.85 so this direction is good. The real good thing is that the maximum error for any of the years decreased significantly to 0.6. Given that the cutoffs are integer numbers and a perfect model could still have an error of 0.5, that is a good result. Still guessing 218 for TX.

I did this same thing for CA and the rise in R2 was dramatic. No real change in the prediction there either, though. Still 222.

378

@WRUAustin LOL! You scored high on Math? You don’t say. Still hoping you are right. Three more months to obsess.

379

@WRUAustin Do you think you can predict DC’s cutoff?

Yr CR M W Total students
08 88 78 74 4370
09 83 91 79 4261
10 87 81 110 4686
11 78 63 103 4197
12 102 75 109 4284
13 76 78 53 4564
14 83 123 79 4438

Thank you!

380

@didierdrogba

I’m sorry, but there is no meaningful correlation of data for DC. The population is low and the spread in the past years cutoff scores is also low. The past 2 years have been at 224. Given the results from the other states, I would guess that 224 is the most likely, but it is not much more than a guess.