Anonymous wrote:As parents, we may be behind the times. Recent article in the Post suggests that ACT test may now be more common than the SAT:

http://www.washingtonpost.com/local/education/sat-usage-declined-in-29-states-over-7-years/2014/03/15/f4504cfc-a5ff-11e3-8466-d34c451760b9_story.html

So, if CA is like the rest of the country and if an equal number of their students qualified though the ACT as the SAT, then 452/2/1922 = 11.8% (below their pop representation), which I know is closer to the truth than 1.3%. Has to be 1600. No other result makes any sense.

I would go "all in" on this conclusion, without even seeing the next card.

As parents, we may be behind the times. Recent article in the Post suggests that ACT test may now be more common than the SAT:

http://www.washingtonpost.com/local/education/sat-usage-declined-in-29-states-over-7-years/2014/03/15/f4504cfc-a5ff-11e3-8466-d34c451760b9_story.html

So, if CA is like the rest of the country and if an equal number of their students qualified though the ACT as the SAT, then 452/2/1922 = 11.8% (below their pop representation), which I know is closer to the truth than 1.3%. Has to be 1600. No other result makes any sense.

Anonymous wrote:

Anonymous wrote:If memory serves me, about 1,500 kids every year score 2400 on the three part SAT. If you exclude the highly subjective and not particularly correlated writing section, as the Presidential Scolars Program does, you End up with well in excess of 3,000 with 1600 on M+R (this is a rough guess and the actual number could be higher). SAT's market share has steadily declined, with close to half of graduating seniors choosing the ACT instead. There is probably some overlap between the two pools, but not much (why would you bother taking the ACT if you got 1600 on M+R SAT?). So, we can probably assume at least 4,000 kids with perfect (not close to perfect, but perfect) board scores (combined tests) as measured by PSP. With CA as large as it is (14-15% of US population) and the Top 40 plus ties rule, it is almost a mathematical certainty that the qualifying SAT score for CA was 1600. It is also highly likely for any populous state, with 1580/1590 as the only other likely outcomes.

I think some of your estimates might be slightly off. In 2014, only 583 students in the entire nation scored a perfect 2400 on the SAT, not 1500. https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2014.pdf

Here is specific data on 2014 CR+M scores (excluding writing): https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-2014.pdf 1,922 scored 1600. To get a pool of 3,000 Presidential Scholar Candidates (the approximate number named), you'd need to include scores of 1580, 1590, and 1600. Admittedly that's a national number, so it does not necessarily prove or disprove what's happening in CA or MD or any other state. But given that 452 Candidates were named from CA, I'm pretty confident that not all scored 1600. If they had, then CA would be supplying roughly 25% of the 1600 scores for the entire nation, which is significantly more than CA's share of the national population (only 12%).

As an aside, it always amazes me how girls score so much lower than boys on the SAT. Girls represent >50% of the SAT test takers, but there are more boys than girls at every score from 1600 all the way down to 1190. It's not until 1180 that girls start to post more scores. And that same pattern applies in both math and reading when you look at each in isolation, although girls catch up to boys faster in reading than in math. It's only in the writing section that they remain roughly even. http://research.collegeboard.org/content/sat-data-tables

I think that you should compare the number of CA qualifying through SAT only (as opposed to all SAT and ACT takers) to the 1,922 and not the entire group of 452 nominees. This may bring the CA supply more in line with overall population representation. Given the demographics of CA, I would be at all surprised if they were overrepresented in the pool of perfect SATs, as well.

Anonymous wrote:PP here. Thanks for the correction on my estimate of 3,000 perfect M+R SATs to the 1,922 that you report. I believe that the right comparison is between the 1,922 and roughly 1/2 of the Presidential Scholar's nominees, because roughly half qualified based upon ACT scores. Let's gross the "perfect board score" up to 3,000 for this analysis to account for the ACT component. And here is why the CA cutoff must be 1600. If you take the Top 40 plus ties methodology that they use, it must mean that the 41st person must have had the same score as the 40th. In the hypothetical case that the 40th person scored 1590 or lower, then that would mean that 39 students or less scored 1600. This would then imply that CA had 39 of the 3,000 "perfect scores" or 1.3% versus 12% of US population. Even if you use the 1,922 pre gross up number, it would only represent 2.0% of recipients. 1600 as a cutoff is the only way to get the math to work.

You make a very persuasive point about how assuming a 1590 cutoff for CA would suggest that CA students account for only 1.3% of the 1600s versus 12% of the population. But isn't the alternative just as surprising?: If we assume a 1600 cutoff for CA, then CA students account for 23.5% of the 1600s versus 12% of the population. Even if we skim some off for the ACT (and perhaps there's good data on ACT results too), those numbers seem off-base.

But in the end, even though either CA scenario seems odd, I suppose one must be correct. I suppose we could consider the population % part of the analysis too, to see if it makes sense. Does CA's 12% of US population strike you as fitting a normal representation of SAT scores? And since we know it likely cannot be that CA is supplying a truly proportional % of the 1600s -- since we know it's either roughly 1.3% or 23.5% -- which appears more likely? I tend to think of CA's top-end students as pretty smart, which might suggest the 23.5% number makes more sense. But on the other hand, I also suspect that a big piece of CA's population is not even taking the SAT, much less scoring 1600, which suggests the 1.3% number might make sense too.

Anonymous wrote:A much simpler skin test can be applied to this. The National Merit Qualifying Score for DC is in fact higher than either VA or MD. It has been for several years. Given the similarity in format with the SAT, PSAT scores are very highly correlated with SATs. The NMSF qualifying score is determined by the top 1% in a given state or district. Even if the pool of DC applicants is smaller (and remember DC while smaller has all of the independent school kids on the Presidential Scholars list, even those residing in VA or MD), we are only talking about the tail end of the distribution. It only takes 40 kids total to determine the cutoff. So, perhaps a 10 point difference, but unlikely much more and the sign could go either way.

I fear this analysis might suffer from at least one false assumption: You assume that DC's higher NMSF cutoff says something about the number of high-scoring students in DC, but you may not appreciate that DC's cutoff is artificially set by NMSC to match the highest state cutoff in the nation, and so does not reflect the score of DC's top 1% of test takers.

Also, to be honest, I am having trouble following your logic. Could you please explain again? In the meantime, I'll get a cup of coffee so my brain will be primed to take it in! TIA.

But DC's list of Presidential Scholars only includes kids who live in DC, while NMSF includes all the MD and VA kids who go to school in DC. I would guess that TJ's presence makes it a lot tougher to make the cut for the VA kids going to school in DC.

Anonymous wrote:

Anonymous wrote:Because the list of Presidential Scholar candidates is purely a matter of having perfect/near perfect SAT/ACT scores. The fact that DC's threshold is considerably lower than DC or Maryland's gives a huge boost to the DC private schools. ...

I'm not so sure you're correct about DC's threshold being lower. I haven't seen much direct data on this point, but you might be able to get a rough sense by looking at the state SAT reports:
CA - https://secure-media.collegeboard.org/digitalServices/pdf/sat/CA_14_03_03_01.pdf
DC - https://secure-media.collegeboard.org/digitalServices/pdf/sat/DC_14_03_03_01.pdf
MD - https://secure-media.collegeboard.org/digitalServices/pdf/sat/MD_14_03_03_01.pdf
VA - https://secure-media.collegeboard.org/digitalServices/pdf/sat/VA_14_03_03_01.pdf

I don't have time right now to dig into all the numbers, but I just did a little quick math to see what % of students in CA and DC scored in the top section (700-800) on the sub-tests. A quick glance suggests about the same % of DC students as CA students are in that top range (although it might bear checking closely because I see some big swings in some of the sub-tests).

Perhaps others have better ideas on the best methodology to approach this problem in light of the limited data we have.

A much simpler skin test can be applied to this. The National Merit Qualifying Score for DC is in fact higher than either VA or MD. It has been for several years. Given the similarity in format with the SAT, PSAT scores are very highly correlated with SATs. The NMSF qualifying score is determined by the top 1% in a given state or district. Even if the pool of DC applicants is smaller (and remember DC while smaller has all of the independent school kids on the Presidential Scholars list, even those residing in VA or MD), we are only talking about the tail end of the distribution. It only takes 40 kids total to determine the cutoff. So, perhaps a 10 point difference, but unlikely much more and the sign could go either way.

Anonymous wrote:

Anonymous wrote:If memory serves me, about 1,500 kids every year score 2400 on the three part SAT. If you exclude the highly subjective and not particularly correlated writing section, as the Presidential Scolars Program does, you End up with well in excess of 3,000 with 1600 on M+R (this is a rough guess and the actual number could be higher). SAT's market share has steadily declined, with close to half of graduating seniors choosing the ACT instead. There is probably some overlap between the two pools, but not much (why would you bother taking the ACT if you got 1600 on M+R SAT?). So, we can probably assume at least 4,000 kids with perfect (not close to perfect, but perfect) board scores (combined tests) as measured by PSP. With CA as large as it is (14-15% of US population) and the Top 40 plus ties rule, it is almost a mathematical certainty that the qualifying SAT score for CA was 1600. It is also highly likely for any populous state, with 1580/1590 as the only other likely outcomes.

I think some of your estimates might be slightly off. In 2014, only 583 students in the entire nation scored a perfect 2400 on the SAT, not 1500. https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2014.pdf

Here is specific data on 2014 CR+M scores (excluding writing): https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-2014.pdf 1,922 scored 1600. To get a pool of 3,000 Presidential Scholar Candidates (the approximate number named), you'd need to include scores of 1580, 1590, and 1600. Admittedly that's a national number, so it does not necessarily prove or disprove what's happening in CA or MD or any other state. But given that 452 Candidates were named from CA, I'm pretty confident that not all scored 1600. If they had, then CA would be supplying roughly 25% of the 1600 scores for the entire nation, which is significantly more than CA's share of the national population (only 12%).

As an aside, it always amazes me how girls score so much lower than boys on the SAT. Girls represent >50% of the SAT test takers, but there are more boys than girls at every score from 1600 all the way down to 1190. It's not until 1180 that girls start to post more scores. And that same pattern applies in both math and reading when you look at each in isolation, although girls catch up to boys faster in reading than in math. It's only in the writing section that they remain roughly even. http://research.collegeboard.org/content/sat-data-tables

PP here. Thanks for the correction on my estimate of 3,000 perfect M+R SATs to the 1,922 that you report. I believe that the right comparison is between the 1,922 and roughly 1/2 of the Presidential Scholar's nominees, because roughly half qualified based upon ACT scores. Let's gross the "perfect board score" up to 3,000 for this analysis to account for the ACT component. And here is why the CA cutoff must be 1600. If you take the Top 40 plus ties methodology that they use, it must mean that the 41st person must have had the same score as the 40th. In the hypothetical case that the 40th person scored 1590 or lower, then that would mean that 39 students or less scored 1600. This would then imply that CA had 39 of the 3,000 "perfect scores" or 1.3% versus 12% of US population. Even if you use the 1,922 pre gross up number, it would only represent 2.0% of recipients. 1600 as a cutoff is the only way to get the math to work.

Anonymous wrote:Because the list of Presidential Scholar candidates is purely a matter of having perfect/near perfect SAT/ACT scores. The fact that DC's threshold is considerably lower than DC or Maryland's gives a huge boost to the DC private schools. ...

I'm not so sure you're correct about DC's threshold being lower. I haven't seen much direct data on this point, but you might be able to get a rough sense by looking at the state SAT reports:
CA - https://secure-media.collegeboard.org/digitalServices/pdf/sat/CA_14_03_03_01.pdf
DC - https://secure-media.collegeboard.org/digitalServices/pdf/sat/DC_14_03_03_01.pdf
MD - https://secure-media.collegeboard.org/digitalServices/pdf/sat/MD_14_03_03_01.pdf
VA - https://secure-media.collegeboard.org/digitalServices/pdf/sat/VA_14_03_03_01.pdf

I don't have time right now to dig into all the numbers, but I just did a little quick math to see what % of students in CA and DC scored in the top section (700-800) on the sub-tests. A quick glance suggests about the same % of DC students as CA students are in that top range (although it might bear checking closely because I see some big swings in some of the sub-tests).

Perhaps others have better ideas on the best methodology to approach this problem in light of the limited data we have.

08:59 -- this is all based on a single test score -- and perfect or close to perfect ones at that. Many kids go on to fantastic colleges without perfect scores. Our oldest child is a senior in a public school and is headed to an Ivy (which some people care way more about than we do, as neither of us went to an Ivy and things turned out just fine). He wasn't a national merit semifinalist or even commended based on his PSAT scores; his SAT scores went way up -- but they weren't close to perfect. Bright kids will bloom where they are planted. I happen to think that for middle and high school the most important thing is that kids are happy and well supported by teachers who really get to know them. (So after talking to him a lot about it, our second child is in private for HS). As long as those things are true, and they are dedicated to working hard, things will work out fine. So enjoy elementary school!

Anonymous wrote:If memory serves me, about 1,500 kids every year score 2400 on the three part SAT. If you exclude the highly subjective and not particularly correlated writing section, as the Presidential Scolars Program does, you End up with well in excess of 3,000 with 1600 on M+R (this is a rough guess and the actual number could be higher). SAT's market share has steadily declined, with close to half of graduating seniors choosing the ACT instead. There is probably some overlap between the two pools, but not much (why would you bother taking the ACT if you got 1600 on M+R SAT?). So, we can probably assume at least 4,000 kids with perfect (not close to perfect, but perfect) board scores (combined tests) as measured by PSP. With CA as large as it is (14-15% of US population) and the Top 40 plus ties rule, it is almost a mathematical certainty that the qualifying SAT score for CA was 1600. It is also highly likely for any populous state, with 1580/1590 as the only other likely outcomes.

I think some of your estimates might be slightly off. In 2014, only 583 students in the entire nation scored a perfect 2400 on the SAT, not 1500. https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-writing-2014.pdf

Here is specific data on 2014 CR+M scores (excluding writing): https://secure-media.collegeboard.org/digitalServices/pdf/sat/sat-percentile-ranks-composite-crit-reading-math-2014.pdf 1,922 scored 1600. To get a pool of 3,000 Presidential Scholar Candidates (the approximate number named), you'd need to include scores of 1580, 1590, and 1600. Admittedly that's a national number, so it does not necessarily prove or disprove what's happening in CA or MD or any other state. But given that 452 Candidates were named from CA, I'm pretty confident that not all scored 1600. If they had, then CA would be supplying roughly 25% of the 1600 scores for the entire nation, which is significantly more than CA's share of the national population (only 12%).

As an aside, it always amazes me how girls score so much lower than boys on the SAT. Girls represent >50% of the SAT test takers, but there are more boys than girls at every score from 1600 all the way down to 1190. It's not until 1180 that girls start to post more scores. And that same pattern applies in both math and reading when you look at each in isolation, although girls catch up to boys faster in reading than in math. It's only in the writing section that they remain roughly even. http://research.collegeboard.org/content/sat-data-tables

Anonymous wrote:I have a first grader at a charter but am constantly thinking about next steps. We like Stoneridge and Visi and Gonzaga and St Anselms for our son. Why aren't these schools on the list? I thought they had top academics? Please educate me!

Because the list of Presidential Scholar candidates is purely a matter of having perfect/near perfect SAT/ACT scores. The fact that DC's threshold is considerably lower than DC or Maryland's gives a huge boost to the DC private schools. The number of students who test at the very top end is not a reflection of the quality of the school. It is more a reflection of how they pick their students. The public magnets are selecting based on test taking ability. Independent schools vary in how they use the WPPSI and SSAT/ISEE. The top scorers and the average scores are very bad indicators of how good a school is. SAT scores are boosted by extensive, expensive prepping. Just because my DS's classmate scores perfectly on the SAT doesn't mean anything about the education DS receives. And, the student with a perfect score may likely have done just as well no matter where school she went to.

I have a first grader at a charter but am constantly thinking about next steps. We like Stoneridge and Visi and Gonzaga and St Anselms for our son. Why aren't these schools on the list? I thought they had top academics? Please educate me!

MD private, sample size of 4 all 1600.

Anonymous wrote:

Anonymous wrote: I think in Md you typically need 1600 -- there are probably more than 30 kids in the state who do that.

Source? For this or the similar California claim above.

NP - I don't think they announce cut off scores like National merit. But my D (MD public) is in with 36 and her friends who made the cut with SAT have 1600.