kitchen table math, the sequel: GPA Calculations

Wednesday, November 24, 2010

GPA Calculations

I was looking at the way my son's school calculates GPA and weighted GPA. Then I looked at how other schools calculate them. Since I was looking at high school web sites, I was only partially successful. In any case, many of them seem to be math-challenged. Even when I searched the web about how colleges calculate GPA (or not), I met with little success. We are told that colleges generally like to recalculate their own GPA. Why? If they have the numeric percentage grades, why do they have to sort them into GPA buckets (e.g. 83-87 is a 'B' = 3.0) and then come up with a new number that contains less information than the percentage grade. Perhaps some high schools send out only GPA bucket scores for each class, but do colleges know what the buckets are?

What I find more interesting are discussions that talk about how colleges like to use unweighted GPA scores. There is some sort of advantage to an 'A' in a regular course over a 'B' in an honors course. Maybe. I think it depends on whether you make it past an initial cut. Someone told me once that colleges use unweighted GPAs for the initial cut. That seems backwards to me. You want to use a weighted GPA (and rank and SAT and etc.) to make the initial cut, but then compare applicants based on unweighted GPA. Of course, they must first strip off all of the non-academic courses.

There is the Ivy League "Academic Index", which uses something called the "Converted Rank Score" (CRS) along with your SAT scores to come up with a number, that presumably, is used to determine whether you make it past a cutoff. The CRS number is based on your class rank and the size of the graduating class. Since most high schools rank based on a weighted GPA (including all fluff courses), this would indicate that class rank matters more than the unweighted GPA of core academic courses. That seems to be a poor game to play. If you are on the bubble, perhaps you might want to look at another college. I told my son that it might be nice to look at his GPA, but he should just try to get the best grades on his core academic courses. This came up because some kids at his high school are getting weird about GPA scores.

I do, however, tell him not to give away free points. For example, in science, they had a grade for whether they covered their books or not. I showed my son that if he didn't do it, his overall grade for the class would drop 3 points. I told him that it was like finding free money. Then there are the things you have no control over, like group grades. There was also the essay where he got an 80, but that was the second highest grade. Both of those things dropped his quarter grade by 5 points. There is not much you can do if a school has two different grading rules for the same course and you stuck with the wrong teacher. Things might average out in the end, but it's difficult while the student is going through it.

82 comments:

gasstationwithoutpumps said...

Different colleges use different schemes. The University of California and California State University both use a GPA calculation that uses an A=4, B=3, ... scheme but adds and extra point for AP and Honors classes, and California high schools have to get their list of honors classes approved. Most of the Calif. high schools use the same GPA computation internally, which simplifies things.

The colleges that use rank rather than GPA do so to try to counteract the enormous grade inflation that has happened in high schools. For example, the principal at my son's school boasted that over half the school was on the honor roll (GPA 3.0 or higher). This despite the fact that only 33% go on to 4-year colleges and only 55% are even eligible (based on courses passed) to enter UC or CSU.

Grade inflation is not the same in all high schools, so a 3.0 average at one school may correspond to a 4.0 at another.

linsee said...

Grade inflation is not the same in all high schools, but class rank isn't either, in that a student's class rank at a high-performing or selective high school will be lower than for the same (or better) performance at a school where many more of the class are low performing.

Bostonian said...

To interpret class rank it helps to know what the average standardized test scores (SAT I and SAT II, ACT) are. Being the median student has different implications when the average SAT score is 2100 vs. 1200.

Politically, standardized-test-adjusted class ranks will not fly because they would penalize black and Hispanic students who attend schools with lower test scores.

I live in an affluent suburb of Boston. One factor tipping me away from sending my children to a well-regarded prep school like Roxbury Latin is that the same academic performance would result in a lower class rank.

Allison said...

--We are told that colleges generally like to recalculate their own GPA. Why?

Because not all colleges want to count PE or ceramics in their GPA calculation, and many don't consider an "honors" class worthy of an extra GPA point unless it's pre-approved.

Some colleges don't take freshman year grades, but do compute cumulative GPA for all terms after that. And colleges have their own value for pluses and minuses.

I'm not quite sure why anyone would think class rank is a better indicator, since it's purely derived from GPA. If you really care you could ask how class rank is reported, since you can expect that for the honors kids, there is probably very little difference between their grades in their courses, and quite common to have 6 or 10 kids with the same computed over-4 GPA.

But colleges know all of this, and that's why schools report gpa quintiles and SAT quintiles too. Colleges aren't confused about how a 4.6 student from school X compares to a 4.6 student from school y. They already know.

Crimson Wife said...

A lot of schools these days don't calculate class rank but rather put students into quintiles or quartiles. The public high school I attended only calculated the class rank for the top 10 students in the grade and then did quintiles for the rest (which really stinks for #11 in the class!) Many prep schools don't do any class ranking at all.

PhysicistDave said...

>I do, however, tell him not to give away free points. For example, in science, they had a grade for whether they covered their books or not. I showed my son that if he didn't do it, his overall grade for the class would drop 3 points.

At that point, I would have urged my son to drop the course and find some academically-oriented course instead.

Dave

ChemProf said...

Dave, I sympathize with the impulse, but you're assuming there is a high school class that DOESN'T grade on that kind of thing.

Lisa said...

Ah, the extra credit points. Ds took cornbread to his Honors English class for extra credit today. ???? Makes me wonder what his A is really worth. My other ds got Honors Geometry points for bringing in Kleenex. Makes me wonder about grades and GPA's in general.

lgm said...

GPA is a joke when the grading policy and extra credit policy varies widely from teacher to teacher. Then add in the extra credits and test retakes given only to students who know to ask on the side.

Also consider that it is a lot easier to get an A/90+ in an honors course than a reg ed course because the honors course is not omitting chunks of the curriculum in the attempt to game the state exam. A reg ed student is assigned more homework and has more independent studying to do to make up for the inefficient use of class time.

If we didn't have end of course state exams and SATs, it would be hard to tell who is getting a high grade for political reasons, who is putting in effort and pulling up the grade with art projects, and who actually learned the material. I'm very thankful for the outside-the-school music judging system also.

I'm in the situation of having a student who'll be in top 1% locally...but w/o AP or IB classes, I doubt he'll be attractive to MIT, Cornell, NYU or the other schools that have admitted vals from here in the past. Can't see CC classes on the resume being anywhere near as impressive as several APs or an IB diploma (we don't have these any more). At this point, I'd rather have him in a high school where he'll have intellectual peers and honors classes & actually be challenged and learning at his instructional level. Better to be #3 with an education than #1 with fluff on easy street.

PhysicistDave said...

Chem Prof wrote to me:
>Dave, I sympathize with the impulse, but you're assuming there is a high school class that DOESN'T grade on that kind of thing.

I was of course being a bit flippant with Steve – he and I have a longstanding debate about whether the system can possibly be fixed. I don’t think it can.

However, I myself did have some high-school classes that eschewed the sort of nonsense he mentions. Shopping around for teachers can work wonders.

I'll admit, though, that my bottom line would be to search for one of those very, very rare schools (lrg claims they do exist; I’m not so sure) that really focus on learning or, alternatively, to homeschool.

By the way, thanks again for the recommendation on the “Asimov on Chemistry” book – it worked very nicely.

Dave

PhysicistDave said...

lgm wrote:
> At this point, I'd rather have him in a high school where he'll have intellectual peers and honors classes & actually be challenged and learning at his instructional level. Better to be #3 with an education than #1 with fluff on easy street.

A few years ago, I was at the in-laws house, and a serious discussion started about whether the high-school transcript looked better with a B in an honors class or an A in a regular class. Finally, my brother-in-law’s sister broke in and asked if maybe we could consider in which class the students actually learned more.

The conversation sort of stopped.

I once earned a guaranteed A in an advanced physics class because I managed to solve one of the prof’s research problems for him – I didn’t need to do any homework, take the midterm or final, or do anything, and I’d still get an A.

I did all the work and took all the exams anyway. I had the quaint idea that I was taking the class to learn something.

Dave

SteveH said...

"Because not all colleges want to count PE or ceramics in their GPA calculation.."

That wasn't my point. It was why would they convert a percent grade to one that has buckets, as I call them? I know that many colleges don't include non-academic classes. I do notice that most 4.0 grading systems give a 4.0 for grades down to a 93 or 94. Does this mean I can tell my son that getting higher grades than that doesn't matter? So, each class is placed in a GPA bucket and then the bucket numbers are averaged. This is quite different than averaging the percent grades and then converting it to a GPA.

"At that point, I would have urged my son to drop the course and find some academically-oriented course instead."

This is an honors science course.


"Ah, the extra credit points."

My son got lots of nice extra credit in honors English for reading a book during the summer and then giving a book report in class. That made up a little for the poor grade he got because all kids in a group project he had got the same grade. I wanted to ask the teacher whether she wanted a group salary based on results.


There are different spins you can put on all of this at KTM. My reason for posting is that I'm going through this now with my son. We know perfectly well what real learning is all about, but that's not always what goes on at his school. We have to play the game. We can't stick our noses in the air and pretend that the carrot is not still dangling in front of us.

Although I tell him to do his best in his core academic courses, we are not going to pretend that the formulas don't matter. In fact, a sensitivity study of the exact formulas can often make it easier to play (ignore?) the game. The fact that he is only one of two freshmen to get an honors boost in orchestra is a big deal. That didn't just fall into his lap. He had to go after it.

Allison said...

-- It was why would they convert a percent grade to one that has buckets, as I call them?

I never received a percent grade on a transcript, and I don't most high schools give them. But assuming there are some, the answer is that the numbers are too precise without having clear meaning (what's a 96 at school X vs. a 94 at school Y), and the buckets are a way of normalizing those differences.

gasstationwithoutpumps said...

-- "It was why would they convert a percent grade to one that has buckets, as I call them?"

I've seen transcripts with percents on them, but only from India. Many US schools don't base the A/B/C distinction on percents, but allow the teachers to assign the grades based on whatever system works best pedagogically.

ChemProf said...

Yes, in American high schools, the percent system is rare, and colleges (especially large schools) are interested in the simplest possible system to compare thousands of students. That means a 4 point GPA calculated in some internally consistent manner, so I'll guarantee they'll recalculate your son's GPA. Basically, they are trying to compare your son to the other students who applied, not particularly consider him on his own merits.

Glen said...

I asked one of the top guys at Stanford's office of undergrad admissions whether they would rather see an A for a regular class or a B for an AP class. He (very firmly) said they wanted to see an A in the AP class. I interpreted this as a warning not to waste time trying to game the system in order to secure a higher spot on the reject pile.

Of course, his other comments made it clear that they had ways of evaluating candidates with a very wide range of backgrounds, so they can and do deal with the question I asked, when necessary. But he also made it pretty clear that, usually, the question was moot, because some other candidate would take the AP class and get an A.

So, should a kid take every AP class offered? No, he said, he should take the most challenging courses he can handle, then prove he can handle them, then his results will be compared to those of other candidates.

So they don't insist that you kill yourself, but if someone else accomplishes more than you do, then you lose. And, not surprisingly, that's the bottom line.

Not great news for the mental or physical health of our kids, but there it is.

Allison said...

--Not great news for the mental or physical health of our kids, but there it is.

Actually, I totally disagree. It is great news to figure out early in life just how hard other people can work relative to you, and whether you can compete with them or not. What's awful is figuring out too late in life.

There are at least several hundred mllion people on this planet who would do our jobs for 1/100th the pay we would do them. There are people in every walk of life who are going to be more talented than we are at something, and maybe at everything than we are. This does not mean we don't have intrinsic worth---it means in fact, we should stop measuring our soul's value against these measures, and then accurately assess who we are.

But if you can't cake walk the 6 AP junior year while taking a sport and a volunteer gig? It's not going to get easier to handle college life. And if the reasons you aren't handling it now are maybe because you have some emotional issues you've not worked out yet, or your immune system can't handle the cortisol your body produces when you work that load? Then you should re-assess your ability to work in pressure cooker environments like med school to be an orthopedic surgeon, or your plan to be an Aero Astro major at MIT.

Character flaws like laziness should be fixed younger rather than older, too.

The ludicrous notion that we can just reach any dream we want because we want it badly hobbles our children. Instead, taking a true stock of our strengths and weaknesses, and then learning to play to our strengths while trying to improve on the margin our discipline, time management or stress handling is the way to go.

Stanford doesn't need to take anyone less than excellent, by whatever measure they value. It's okay not to meet their measure, but it's far more important not to lie to oneself about what those measures are. If you really don't have the ability to ace your high school AP courses with ease, or you lack the drive to ensure you ace them anyway, you aren't going to go to Stanford.

ChemProf said...

At Stanford, their entering class has the following profile:

SAT reading over 700: 62%
SAT math over 700: 73%
SAT writing over 700: 69%
GPA 3.75 or higher: 91%

So yeah, a couple of B's won't eliminate your chances, but fundamentally, they want to see A's. When you are admitting 2300 students out of a pool of 32,000, you get to be that picky! And when you are doing that, you want to make it as easy to compare students as possible.

All this info is from the Common Data Set, by the way, and you can find similar info for any target school
http://ucomm.stanford.edu/cds/2010.html

Bostonian said...

I think Allison is exaggerating a bit for effect when she says a student should "cakewalk" 6 AP classes in one year to be able to handle the academic pressures at a selective college. About 20 years I took 4 AP exams (getting 5's) and did fine at an Ivy. Whether one now needs lots of AP courses to GET IN to an Ivy is a different question.

Students accepted to Ivies may be bright, but they are also ignorant of many basic things, as shown by the C-grade performance on an American civic quiz: http://www.americancivicliteracy.org/resources/quiz.aspx . One should not be over-awed by them.

Allison said...

My comment said nothing about their *knowledge*. I was talking about their ability to handle coursework, sports, hobbies, and life in general.

A student who is debating "should I take this AP course when I lack confidence I can get an A, or should I take this reg course where I can get an A" is not likely Stanford material, because there are plenty of student who have complete confidence they can, and they are right. And yes, for many, many students, taking 6 APs in a year just isn't hard. They are busy, but they aren't struggling. Some people thrive on that. Others do not. For the ones that do, bluntly, it's just not that difficult for them to get As in AP courses, or 5s on AP tests, and it's not difficult for them to play a sport, or volunteer, or be on the debate team, or any of those other resume-filling things kids getting into Stanford have.

Again, this says nothing of *the knowledge* of these students, or the education they received in such. But if a student really is having this debate because they doubt their ability to succeed with As in AP courses, then somewhere they are lacking the aptitude, preparation, or drive to compete with the top 10000 other candidates applying to the top schools. That doesn't mean they aren't valuable people, or won't have a fulfilling experience in college, or several successful careers. But it does mean they aren't likely to get into e.g. Stanford, and should spend more time figuring out what a good fit is than worrying about the hundredths position of their GPA.

Crimson Wife said...

I'm a Stanford grad, and the only folks who are getting in with <3.75 GPA or <700 on any of the SAT sections are those who are in their own admissions pool (recruited athletes, legacies, scions of the uber-wealthy aka "development admits", certain racial/ethnic minorities, etc.) If you're a normal, middle-class, non-legacy, suburban Caucasian or Asian-American kid, you'd better have a sky high GPA in rigorous courses and SAT score.

Lisa said...

All of which is interesting but if your kid isn't interested in the Ivy League what do 'regular' colleges think? Grades aren't about being accepted for us as much as they are about scholarship $$$ frankly. Oh, and I wanted to mention our high school uses an 11.0 GPA scale, just to complicate things.

SteveH said...

"...in American high schools, the percent system is rare, "

It is????

Then surely they must use +/- on the transcripts. I can't imagine they send out a GPA number for each class unless it is resolved to a +/- level. It seems to me that colleges would want the raw data so they can decide how to truncate (in)significant digits rather than letting the high school do it.


"...and the buckets are a way of normalizing those differences."

I can see how one might claim that there is no difference between a 94 and a 96, but I don't see how it "normalizes" the differences between schools. Besides, most GPA buckets lump everything together for grades between 93 and 100 as a 4.0 and the cutoff at the low end varies at most schools. Is there really no difference between a grade of 99 and a grade of 93? This is important. I will have to find our exactly what our high school sends out on their transcripts. My son gets percents on his report card, and each of his classes is fairly careful about weightings for grades.


This is not so much about gaming the system as it about allocating resources. It has to do with not getting caught up in all of the angst generated by other students. One way to do this, I suppose, is to ignore it all and just focus on doing your best in you academic courses. I think that we do both; know the details and keep it all in perspective.

So, from what I can tell, the goal is to get work on classes to get them into the next higher GPA bucket. The key one seems to be the huge 4.0 bucket.

SteveH said...

"Grades aren't about being accepted for us as much as they are about scholarship $$$ frankly."

Good point. My brother-in-law talked about how numbers in formulas mattered when it came to a merit scholarship for his daughter. It translated into $$$. It would be nice to see some of those formulas. In her case, I think her ACT score was huge and her GPA was used as only support information. (One college even suggested that she could retake the ACT to see if she could raise her score. She had taken it only once. Actually, she got really pissed off at the comment and dropped the college from her list.) The fact that she got a 34 on the ACT was huge; much more so than how many AP classes she took. Actually, she was in an IB program, but I'm not sure how colleges feel about the IB program.

This also raises the issue of how much you can prepare for the ACT exam versus the SAT exam. Is it possible to improve an ACT score more than a SAT score by study?

Allison said...

-- It seems to me that colleges would want the raw data so they can decide how to truncate (in)significant digits rather than letting the high school do it.

Steve, a composite grade in a semester long course *isn't* raw data. Raw data would be the tests and homework, conversations in class, combined with the syllabus for grading scale, etc. Colleges never see the "raw data". They couldn't handle that level of data anyway.

You get worked up about how different teachers might have different grading systems at one school, but teachers at thousands of different schools have thousands of different grading systems, so in the end, there isn't really an apples-to-apples comparison unless you're rounding, and that rounding means a simplified GPA system. Schools have a bucket notion of an A because that's about the accuracy of the grading systems employed around the country.

Sure, some computations take into account the pluses and minuses, but honestly, it's just a proxy for student performance anyway, and a piece of the application.

The top students all hit the ceiling on GPA anyway, and that's why more digits of precision isn't providing more accuracy.

For scholarships, this is the same. If you're busy worrying about the formula, you're missing that there is already some other student who out-scored yours on GPA and SAT, etc. Someone hit the high without spending time gaming it. For the most prestigious scholarships, this means you need the top scores to get it. For less prestigious ones, the rest of your package matters--your personal story, your recommendations, your connection to the scholarship/school/focus.

Time is better spent taking the hardest courses you can and doing your best in them than worrying about how the numbers get computed.

(Another good lesson would be to not let teenagers make teenager mistakes, like getting piqued and punting a school for telling her to retake a test. They were giving her good information, and she acted irrationally.)

gasstationwithoutpumps said...

Actually, retaking a test is not necessarily good advice if you already have a high score. There is a certain amount of random noise in any test score, and if the score is better than you would expect from other evidence of achievement, then it may have been inflated by luck. In that case, retaking the test is likely to reduce, rather than increase the score (a phenomenon know as "reversion to the mean").

Neither SAT nor ACT scores benefit much from specialized studying, for students who already know most of the material on the test. The reported boosts from studying for the standardized tests mainly come from those who did not learn the material in the first place, and scored poorly. The College Board has some tech report on test-retest results, and while there is a boost in the middle and at the bottom, at the top the reversion to mean wipes that out. (Sorry, I don't have a pointer to the report, and it might take me a long time to find it on their reorganized web site.)

Allison said...

When a school tells you to retake a test, they are telling you something: your score is not high enough. That was good information.


Yes, her score might have gone down (though my understanding is you can choose to have only your highest score reported) but it wasn't high enough in the first place or they wouldn't have told her that, so she had nothing to lose.

ChemProf said...

For scholarships, there isn't a general formula. Top ranked schools (Ivies or top 25 liberal arts colleges) don't give much merit aid because they don't have too. State schools also don't give a lot, although there is a little depending on the school. If you really want to maximize your scholarship dollars, look at small liberal arts colleges that aren't in the top 25 (say those ranked 26-75 on US News and World Report). These schools want to attract strong students to keep their ratings up, and will use merit aid to do so. However, you can't make a prediction ahead of time. I can tell you that at my tiny college, anyone with average SATs above 650 (especially over 650 math) and a GPA of 3.5 or better will get some amount of merit aid, but that can range from a $5000 scholarship up to a free ride, depending on the student and how excited admissions is by them.

Crimson Wife said...

The reported boosts from studying for the standardized tests mainly come from those who did not learn the material in the first place, and scored poorly. The College Board has some tech report on test-retest results, and while there is a boost in the middle and at the bottom, at the top the reversion to mean wipes that out.

I don't believe that for one second. I know a *TON* of folks who raised their SAT score by 100+ points through test prep who already started out well above average. I was one (1250 equivalent on PSAT to 1350 on SAT attempt #1 to 1450 on SAT attempt #2, all pre-recentering).

Bostonian said...

Crimson Wife wrote "I know a *TON* of folks who raised their SAT score by 100+ points through test prep". The study Preparation for College Admission Exams found an average improvement in SAT scores (on the 1600 scale I think) of about 30 points from test prep.

The SAT is largely an intelligence test, so test prep cannot improve scores much. The College Board must strenuously deny that the SAT is an intelligence test because of the approximately 1 SD test score differential between whites and blacks (of a similar magnitude to the differential on IQ tests).

Crimson Wife said...

The SAT absolutely can be prepped. Kaplan and Princeton Review wouldn't be making millions of dollars if it wasn't. Taking standardized tests is like any skill- the more one practices, the better one gets at doing it.

The College Board doesn't want to admit how big a factor test prep plays into SAT scores because doing so would give ammunition to the critics of the test.

gasstationwithoutpumps said...

"Kaplan and Princeton Review wouldn't be making millions of dollars if it wasn't."

This is not a convincing argument. People have made millions selling snake oil before.

http://www.collegeboard.com/about/news_info/sat/satfaqs.html#testprep
gives the College Board take on test prep, and points to research (not by College Board) supporting their view that short-term test prep doesn't help much:
http://www.nacacnet.org/PublicationsResources/Research/Documents/TestPrepDiscussionPaper.pdf

"This consensus is as follows:
• Coaching has a positive effect on SAT performance, but the magnitude of the effect is small.
• The effect of coaching is larger on the math section of the exam (10–20 points) than it is for the critical reading section (5–10 points).
• There is mixed evidence with respect to the effect of coaching on ACT performance. Only two studies have been conducted. The most recent evidence indicates that only private tutoring has a small effect of .4 points on the math section of the exam."

SteveH said...

"Steve, a composite grade in a semester long course *isn't* raw data."

Allison, can't you see what I'm talking about or are you bound and determined to spin it your own way?

"Schools have a bucket notion of an A because that's about the accuracy of the grading systems employed around the country."

So, you're saying that that even though the information is there, colleges don't care about the differences between a grade of 93 and a grade of 100?

"Time is better spent taking the hardest courses you can and doing your best in them than worrying about how the numbers get computed."

The fact that colleges don't see any difference between a 93 and a 100 doesn't tell me anything useful? My son is already taking the hardest courses he can, thank you. If he has a class that is on the borderline between the A and A- break, it tells him a lot, even though his his total percent grade average is a 97. Thanks for your opinion, but I'm looking for facts. I'm perfectly capable of making my own judgment about what is important.

palisadesk said...

Crimson Wife and the SAT research could both be correct. The data report group averages, whereas individuals might have widely varying outcomes. Yes, the SAT is closely related to IQ (and domain knowledge), but individuals with the requisite IQ and skills might underperform on the SAT or other standardized tests due to poor test-taking skills and strategies, slow reading speed, or personal quirks such as anxiety levels or perfectionism. Tutoring or test-prep might conceivably assist such individuals in raising their test scores significantly by enabling them to actually reflect their true IQ and skill levels.

I never took any test prep, but did spectacularly well on both the SAT and the GRE (multiple 800's pre-recentering), and while it would be gratifying to think I am a genius, the reality is that my test-taking strategies, not specifically taught, so far as I recall, probably made a significant difference. I set myself time parameters for each section, rapidly worked through each one in the time allowed (or less), first working the, to me, "easy" parts and making a pencil tick in the margin of any item to which I wished to return and reconsider, or which I omitted to the end (no, you are not allowed to make "stray marks" -- I erased those ticks before handing in the test). I used the time left over to do the items I had initially skipped, and to carefully re-examine ones of which I was not 100% sure. This way I completed all sections of the test every time I took it -- a significant factor in my favor I'm sure. I also had ample time left to double-check that I had completed every question to my satisfaction.

Now I am sure there are more sophisticated test-taking strategies than what I used, but students probably quite as intelligent as I was did much less well, and I bet those are the students who could conceivably make 100-point gains on retake. That number would be small relative to the total number who take Kaplan or Princeton Review classes, but these outliers would be sufficient to inspire others to try to do the same. Also, there would be numerous cases where a relatively small score increase (10-20 points) would be sufficient to get the student over some benchmark -- 1200 total, or 1400 total, in the old two-part sum system -- to be competitive for a particular college.

The group mean would be dragged down by the scores of relatively low-ability students who actually do worse when they apply too much reflection to their answers. I have frequently seen this when doing educational assessments on low-ability students: their initial response to an item will be confident and correct, but then they second-guess themselves and change it to an incorrect answer. If they do this more than a couple of times -- and many do -- this drags down their total score significantly. It can even drop them into a lower stanine.

SteveH said...

"The SAT absolutely can be prepped."

I agree. Are people saying that students shouldn't prepare? Statistics might say that it doesn't help much for some mythical average child, but my son is not an average statistic. There is a probablity that it could help him quite a bit, or enough to get him into the next higher bucket. I'm looking for data and facts to help me understand the tradeoffs.

When I look at SAT math questions, I don't see intelligence, I see preparation. Besides, nobody is talking about the amount or kind of preparation. How many hours are they talking about? How long is a typical SAT prep class? I once taught an after-school SSAT prep class and of course it wouldn't help much. It was too little, too late, and the students were just going through the motions. How do you separate the effects of good preparation from bad preparation?

Anonymous said...

If the SAT/ACT tests are IQ tests, then I assume my son's 7th grade scores will be the same as when he takes it junior year, as any IQ test would be. I'm pretty sure that won't be the case.

Of course IQ is a huge factor. Do we need a whole thread for that?

It's an achievement test for the most part. You have to be taught a good deal of what's in the tests. The essay portion of both the SAT and the ACT are highly formulaic, but if you prep for the SAT essay like it's the ACT essay, you could blow it big time.

Stamina plays another part. The SAT is longer. The gifted immature kid with ADHD is probably going to have more trouble with time and focus than some of his other classmates. Time strategies made a huge difference in my kid's scores from one testing to the next.

I only used a couple of prep books, but I noticed the improvement as we went through the answers and then tried it again. I used a tutor for some of the math because the ACT has a few trig questions and my son was only in algebra 1 at the time.

The first test revealed that he missed several easier pre-algebra/algebra questions that were clearly dumb mistakes. However, he managed to get the harder trig ones. The second time he took it in the spring he made no dumb mistakes on the easy ones, but was able to answer the tougher trig questions, landing him in the 99th percentile of all the Midwest Talent Search takers.

In the science portion of the ACT, my 7th grader about had a heart attack at the big words that he had never seen before. Once I told him to replace the big words with "liquid A" or "liquid B", he would see that most of the test was about logic and reasoning. He was then able to score high enough to be accelerated in science, also. I'm glad I didn't send him in there cold.

SusanS

Bostonian said...

SteveH wrote, "I agree. Are people saying that students shouldn't prepare? Statistics might say that it doesn't help much for some mythical average child, but my son is not an average statistic."

In general, when making decisions, one should weigh costs (in this case the cost of a prep course and the time spent by the student) vs. the expected benefit. I provided a title and GSWP provided a link of a meta-analysis of the effect of prep courses, and for some reason Steve wants to ignore it. Does he disagree with the cited 30-point average gain? If so, what does he think the expected gain is, and based on what evidence?

I will prepare my children (currently in early elementary school) for the math SAT by presenting all the topics years before 11th grade, when the score really counts. The eldest has taken the SSAT (used by the Johns Hopkins talent search), and they will take the SAT in 7th grade (when talent searches use it to qualify for middle-school and high school summer gifted programs). Standardized tests in general and the SAT in particular should be "old hat" for them before 11th grade. They use EPGY and Singapore Math and participate in contests such as the Math Olympiad. If they have been working steadily on math for years, a few-month Kaplan prep course is unlikely to make a substantial difference in their scores. I will buy the $20 Kaplan and Princeton review test books and software.

About 20 years ago I scored 1560 on the SAT using approximately this approach, without taking any prep class.

Bostonian said...

SusanS wrote, "If the SAT/ACT tests are IQ tests, then I assume my son's 7th grade scores will be the same as when he takes it junior year, as any IQ test would be. I'm pretty sure that won't be the case."

Your son's SAT scores will likely rise, because he is getting smarter (his brain is still developing) and because he is learning more math. But if all 7th graders took the SAT and later took it in 11th grade, the correlation between 7th grade and 11th grade scores would be highly positive, because smart 7th graders become smart 11th graders.

There is research showing how SAT scores of normal-age test-takers can be equated to IQ scores. It would be interesting to map the scores of early test-takers to IQ.

Yes, the SAT I is to some extent an achievement test, but the sad fact is that most people simply don't have the brains to learn (for example) the required SAT math material and know to apply it quickly and correctly enough to earn a 700.

SteveH said...

Bostonian wrote:

"The SAT is largely an intelligence test, so test prep cannot improve scores much."

You need to define exactly what you mean by "test prep". You seem to have a very narrow definition.


"...and they will take the SAT in 7th grade ..."

This is not a form of "test prep"? From your description, I see a lot of test preparation. Why don't you just leave it up to their natural "intelligence"?

palisadesk said...

I assume my son's 7th grade scores will be the same...., as any IQ test would be.

That is not at all necessarily the case. IQ scores do change, sometimes significantly, in children (even adults). This is more true the farther you go from the mean. IQ scores are most stable in the average range (25th-75th percentiles) while the greatest intra-individual discrepancies are likely in individuals scoring near one extreme or the other. A child who scores in the Developmentally Disabled range at age 7 may score in the Mildly Delayed range at age 10 and in the average range at 14 (I had a student who did this) -- a difference of more than two standard deviations. It is a known trend that students who do not learn to read, or who are very poor readers, will usually show a gradual drop in IQ from early elementary through secondary school; where "learning disability" is defined using the "discrepancy" definition (significant discrepancy between achievement and measured ability), this can be used to reclassify the child as no longer eligible for services. A drop of 15-25 points is not unusual.

IQ tests correlate highly with each other, but not precisely. Thus a child may do much better on the WISC than on the WJ-III, or the other way around. Some people who do very well on Weschler or Stanford-Binet measures do much less well on the Ravens (which is very demanding of visual perceptual reasoning). The notion of a single-point "fixed" IQ does not stand up to empirical evaluation. What is more likely is that "intelligence" (however defined) is a genetically-potentiated range which is affected during life by numerous factors, so that an individual may grow-- or decline -- towards one endpoint of his/her range or the other, in a similar manner to realization of genetic potential for stature or athletic prowess.

The present-day SAT is less closely linked to IQ than it was pre-recentering. I understand that the high-IQ societies no longer accept SAT scores for membership applicants. But the correlation, though looser, is still there.

Allison said...

Steve,

Admissions folks will distinguish between the hundreds and maybe thousands of 4.6667 GPAs by SAT scores, letters of recommendation, and personal essay. Letters of recommendation are where teachers give rankings to students, comparing them against their in-year and recent-year peers. That is how they know who the 98 average student is vs the 93 average student.

Allison said...

The SAT Math questions are extremely good examples of mathematical maturity. I don't know what Steve meant by "When I look at SAT math questions, I don't see intelligence, I see preparation", but the mathematically mature student is going to do extremely well, and no prep course is going to save the mathematically immature.

The questions Catherine posted a bit ago are excellent examples--that one with two points, R and S, as the center of circles, and then the highlighted curve was a semicircle around R and another around S, where you were given the total distance of RS, and asked to find the arc length, another where you were given a triangle of small triangles, told they were congruent, told area of some, and asked to find the area of the large. You needed to see that you didn't need to solve for r and s to solve the problem--and the simplest way to see that was just write down what you were asked in terms of what you were given. then the aha would come immediately, if you were mathematically mature (or would have without writing it down).

These were both problems where the problem solution was IMMEDIATE to someone who read the problem and had maturity. The list of things you had to have unconsciously/subconsciously mastered included: definition of a circle, relationship of RS to radii, definition of arc length, distributive property. For the other, you needed to know: equilateral, congruent, relationship of congruent pieces of congruent objects, how to properly set up a proportion, how to solve proportion. To get a 700 or above on a test like that, you need to be solving all of the questions properly as fast as you can read them. You won't get over a 600 if you can't do them in on average, under 30 seconds.

The typical student who has been getting As in Everyday Math and then is flung into rigorous high school math isn't prepared to get a 600 on such a test. (and I know 5 such students now who are in seniors in AP stats, having straight As in math always, who scored 500.) That's because they don't have that fluency to know what those concepts mean when they read them, they don't have the discipline to turn the problem into a mathematical statement instantly, and they don't really know what is being asked fast enough. In fact, spending more time on those problems really isn't going to help, because there's nothing to work out--either you see it or you don't.

Certainly test prep can help some, but it isn't going to fix deficiencies that are cumulatively created by years of not being taught math concepts to mastery.

Anonymous said...

Palisdesk,

I understand that it is not always the case. I am very familiar with IQ tests and clustering (or not) of scores and subset scores. I have a special needs son who has been all over the map in terms of IQ, so I understand that some kids do not have consistent scores. I'd go into detail, but I already have over the years.

But, as an example, his 5th grade Weschler (not WJ-III, I believe) jumped 20 points from his Kindergarten one (probably the addition of medication helped), yet still contained subset scores with over 9 point spreads indicating an LD(s) or other problems. By the end of high school it had dropped back down to the early grade school score, which was confusing to me, but as I understand, the way things often tend to be with kids like my son.

My other son, however, has always clustered pretty closely in the gifted range on all tests, including on the Ravens, so I'm assuming something new won't be popping up anytime soon, although I can always be surprised.

Don't get me started on the whole special services business. I'm in the middle of that right now and it's unbelievably frustrating. We're going to have to hire a lawyer with a specialty in this area to make sure my son gets help if he needs it.

Apparently, in my state an IQ of 69 means you qualify for assistance (reduce bus fees, Medicaid, SSI), because of the assumption that you cannot get any kind of gainful employment, but one point over means you'll be fine. As you know, the subset numbers can reach way own below 69, but the overall score can be high enough so that your child doesn't qualify for any assistance in any number of agencies that exist for just that purpose. While the adult with an overall IQ of 70 can no more get gainful employment than the adult with the 69 IQ, there is no tiered response from the government. You can or you can't, by one point. Anyway, I don't mean to hijack the thread here. This is just my own personal anecdote.

There really is a simple remedy, though. My husband and I just need to live forever and it'll all be fine.

SusanS

Bostonian said...

Allison wrote, "The typical student who has been getting As in Everyday Math and then is flung into rigorous high school math isn't prepared to get a 600 on [a test such as the math SAT]."

I don't like EM either (our school uses it) and use other curricula with my children at home, but EM is used by most public school systems in Massachusetts, and MA students consistently score among the highest in the country on tests such as the SAT and NAEP. I certainly don't think EM causes this outperformance, but I cannot cite evidence that it is hurting. Is there evidence that Singapore Math or other curricula produce higher math achievement than Singapore?

Allison said...

Bostonian,

A plurality of primary public school systems in the US use EM, and the top 30 by pop all do--why does MA score the highest in the country, when nearly everyone is teaching the same thing?

I mean, this is your bailiwick, right? Your repeated argument is IQ accounts for much of the variation in these results. Surely the home to MIT, Lincoln, Draper, Harvard, BU, BC, Wellesley, and a hundred more universities, plus the 128 belt of high tech corps would have higher IQ than random state in the US, right?
But assuredly, parents in MA spend more money on test prep for the SAT than LA does, or FL does, if it does something, then you'd expect to see it there.

I have seen no comprehensive data on Singapore Math in the US. I have seen little comprehensive data on any of the curricula. But the TIMSS scores are clearly an indication that Singapore Math produces higher math achievement than Everyday Math.

Anonymous said...

EM comes from IL, so very quickly most suburbs and towns, particularly the wealthier ones, started adopting it (or one of the carbon copy versions) several years ago. Interestingly enough, the state tests started including a lot of topics EM covered that others didn't (baby stats, writing explanations instead of using math to solve, etc.) at around the same time as the curriculums came out.

One of the first things I noticed in our district after a couple of years was that the middle school dropped the 6th grade pre-algebra class and just made it a general math class with some pre-alebra, that is, if they kids were ready.

Over the years, several towns and suburbs have quietly dropped EM (according to the site IL Loop), although they don't usually advertise why. Many seem to be going back to a more traditional curriculum, however, some have kept EM and just supplemented with something else.

Our district has been tossing around Singapore of late as it starts to phase out its EM copycat, but like a lot of towns, the fear is that the average grade school teacher it too weak in math and will not have support to teach it properly. Our grade school teachers are not specialists, so until Singapore addresses that, most schools are going to pass, which is a shame.

SusanS

VickyS said...

In the middle of college applications at our household, so will chime in on the grade point average question.

First, on the Common Application, you must state whether your GPA is weighted or not. Whether colleges use this information is up to them.

Second, class rank is important and can help sort out the different GPAs. I wouldn't be surprised if some colleges use a formula that attempts to normalize or adjust GPAs using class rank. Some private schools, like ours, don't report class rank but report in quartiles or some such. This is so that kids who get A's in "survey" or less rigorous classes don't show higher class rank than AP kids. The AP bump (which in our school I think is 5 percentage points, and only if you also take the AP exam) does not really make up for that.

Our school is notorious for grading hard. My son, who is one of the top students, has a GPA of only 3.65 and that includes a bump up for AP classes and A's in fitness class several times over. I was shocked to see, on the college profiles we've been viewing, how the GPAs of the admitees compare with their overall SAT/ACT scores. Sure seems to be evidence of rampant GPA inflation. Given my son's test scores, he should have a 3.8 GPA or higher, but I'm not sure anyone does at his school. All I can guess is that colleges must be able to deal with this, some way or another, because there is no uniform grading in high schools so the best measure of comparison needs to be test scores and perhaps rigor of courses.

Anyway, what the colleges receive from the high school along with the transcript is the high school's "School Profile" which explains how the school grades, how class rank is determined, how weighting is done, average test scores for the last several years' classes, etc. I'm guessing this is a pretty important document when it comes to assessing any individual student. Most schools have this available to the public on their websites.

In our school, here is our grading system. It was recently changed from an even tougher one in which 92 and below was a B.

A+ = 97 - 100% = 4.0
A = 94 - 96% = 4.0
A- = 90 - 93% = 3.7
B+ = 87 - 89% = 3.3
B = 84 - 86% = 3.0
B- = 80 - 83% = 2.7
C+ = 77 - 79% = 2.3
C = 74 - 76% = 2.0
C- = 70 - 73% = 1.7
D+ = 67 - 69% = 1.3
D = 64 - 66% = 1.0
D- = 60 - 63% = 0.7
F = 0 - 59% = 0.0

I teach there too, now, and for each assignment I enter numerical grades into our grading system (Gradebook Wizard). The system automatically converts the grade to a percentage (16/20 would give you 80%) and then assigns a grade (B-). Grades can be weighted by categories if desired (e.g., homework, tests, etc.). By the time it bubbles up to the top, the parent and student will see both a percentage grade and a letter grade automatically generated for each assignment. At the end of the year, only the letter grades are put on the transcript, not the percentages. Then, down the road, when the transcript is sent to colleges, it is accompanied by the school profile, which shows how the percentages relate to the letter grades.

Overall I agree with Allison--it's tough to game the system because it's just too loosy-goosey to begin with, and trying will make you (parent and child) go crazy. In the end, my son and I both agreed that he should be attending the college that wants him "as is" not one that he is trying to game himself into. He is a smart boy with lots of potential, and he is making a good effort, but many smart, hard-working boys drop out of their first year of college because they have gotten in over their heads, have no parent to watch over them, or are otherwise not able to deal with the cold water shock of the real academic demands made on them.

VickyS said...

And one final note as to Steve's "buckets." In college, my transcript only showed A, B, C or D (no pluses, minuses, etc.). Talk about wide buckets! But what really is wrong with that? You could get a 90 or a 99 and still get an A. I think that's okay. Why slice it thinner than that? So I'm all for buckets in high school, too. I think that would ease a lot of the temptation to game. You do your best, cross over that 90% mark, and you get your A. The kids who get A's are the ones that know the material well. They can then be differentiaed on other bases, such as test scores and activities.

ChemProf said...

One issue is that there isn't any one system for college admissions - every school is (very) different. A few examples, that I'm sure have changed in the last 20 years.

When I was at UC Berkeley in the early 90's, they'd just had an admissions scandal involving not admitting qualified Asian students, so they published their admissions grid. They turned the GPA (on a 4 point scale) into a number going up to 4000, then added your SAT score (limited to 1600 at the time), and plugged you into the grid. Above a certain total, you were in, in another range, they might look at your essay. These ranges were, at the time, explicitly set based on your race, but it was very mechanical. Out of more than 10,000 applicants for first year (not transfer) slots, they'd only really read the applications of 1000 or so.

At the other end was Harvey Mudd College. There, they wanted an entering class of ~150 students (it is more now), which meant admitting ~250 from ~1000 applicants. They split the applicants into three catetories: top 100 by the numbers, qualified, and not qualified. The top 100 were admitted, and the next 150 were chosen from the "qualified" applicants, which was about half of the applicants. There, they looked at extracurriculars, letters, and the rest of it. They read a lot of applications very carefully, but didn't have nearly as many to deal with.

So, if you are concerned, you need to be talking to the specific colleges your son is thinking about applying to. It is going to vary a lot!

Allison said...

There are many articles about how admissions departments work. Here is a famous one from a while back, a decade ago. The numbers are changing, but the overall process isn't.

http://query.nytimes.com/gst/fullpage.html?res=9904EEDF1439F934A15751C0A9669C8B63&sec=&spon=&&scp=19&sq=steinberg%20wesleyan&st=cse

This one is more recent, but again, shows the non mechanistic side of admissions:

http://articles.latimes.com/2009/apr/03/opinion/oe-perez3


NYT has a blog with a lot of info on it, called the choice:

http://thechoice.blogs.nytimes.com/

Crimson Wife said...

I don't know whether the school I attended growing up in MA uses EM but I know for a fact that as of 2003 when my youngest brother was a senior, nearly all the students either took a SAT prep course or hired a pricey individual tutor.

SteveH said...

"That is how they know who the 98 average student is vs the 93 average student."

What I'm saying is that the 93 cutoff is a big deal. You lose all finer resolution and colleges have to figure it out some other way. It also affects how my son should allocate his resources. It might all be fine if colleges look at the entire picture for every application, but it's meaningful if GPA places you in a lower comparison bucket. In any case, I sent an email to the guidance department to find out eactly what information they send to each college. All else being the same, my son needs to worry more about classes that are on the borderline. A consistent effort in this area could have a worthwhile payoff.

SteveH said...

"...I don't know what Steve meant by "When I look at SAT math questions, I don't see intelligence, I see preparation","

I see questions that my son has never seen before. It's not maturity that will fix that. It's exposure and practice.


"Certainly test prep can help some, but it isn't going to fix deficiencies that are cumulatively created by years of not being taught math concepts to mastery."

I don't think anyone here thinks that last minute test prep will solve anything. I'm more interested in information that refers to test prep that starts year(s) ahead. Calling it maturity doesn't help me much. All I can do is to look at as many sample questions I can and to classify them for my son. Books like the Art of Problem Solving are good, but, given a limited amount of time, and the desire of my son to focus on increasing his grades by a few points (which may be meaningless), I would prefer a more targeted test prep approach. I'm finding, however, that "test prep" implies not real learning for many people. I don't have that view, and would like to hear what others have done to deal with long term targeted test preparation.

I think that speed is huge, not only with calculation of the basics, but in terms of whether you have seen that type of problem before. My son tends to be slow and methodical, but he makes very few mistakes. There is also the issue of flexible knowledge and how that can be improved. It's not exactly maturity. It's whether he can approach a problem in several ways and get the same result.

lgm said...

Our high school guidance dept tells students to maintain a 95 or greater average in core academic classes if private college is the target. 90 average in core academics for state college. This rec included the weighted average which had 5% boosts for AP, IB, and honors classes. The recommendation hasn't been changed even though AP & IB have been eliminated. All should be pursuing the Regents Advanced Diploma and acheiving the Honors designation (HD is 90 or greater average on all end of course Regents exams). All grades are numerical. No AP available through the school, no decision yet on how an independent study AP will be weighted (this is the first year without AP classes); in any case only 5 can be transferred in.

So yes, a 98 puts one in a different league than a 93 here. And it starts in 8th grade ...one needs decent study skills before one starts high school courses if one is going to be in the top band. Even a math brain needs those skills...seperating all the 99s going in to the midterm is done by giving one problem of a certain type in class, then making that concept worth 25% of the midterm. A child that thinks he's on top of the world with a 99 average going in will be taken down a notch if he didn't catch that one problem during his study & review time. Study skills are a big deal.

SteveH said...

"...it's tough to game the system because it's just too loosy-goosey to begin with, and trying will make you (parent and child) go crazy."

Actually, I'm trying to use the information to avoid having him go crazy. From what I can tell, above a 93 or so, the extra effort (stress) is not needed and he can focus on learning, not trying to get every last point on his tests.

Also, we don't want to be the last to know about extra credit points floating around. The fact that he missed out would cause stress. We are not so proud, and nobody would ever think he was the sort of student who just wants to game the system.

I mentioned before that he is only one of two freshmen who will be getting the honors weight bump for orchestra. He might have done all of the needed performances, but if he didn't ask, he wouldn't have gotten the boost. If he saw others getting the boost and not him, that would have caused a lot of stress.

Now I'm trying to figure out how to focus some of his attention away (long term) from his high school classes to preparation for the PSAT and SAT. If I can show him how those grades are weighted compared to GPA, it might get his attention.

This might sound like I'm setting him up for lots of stress, but it all depends on how it's done. I don't think it would work just to have him do lots of practice questions. I have to do something about organizing and classifying the problems.

SteveH said...

"So yes, a 98 puts one in a different league than a 93 here."

But what numbers actually get sent to colleges on the transcript? That's what I'm asking our guidance department. There are other numbers which could be influenced. In our school, the weighted ranking never goes into a GPA bucket. They take the percent grade times the weight for the class (3 for regular classes, 3.4 for honors, and 3.7 for AP) times the number of credits for the course. As far as I know, that's what gets put on the transcript along with unweighted GPAs. However, they might just send out the class rank and the size of the senior class. If I know exactly what gets sent out by our high school, I can give my son better advice over what is important and what isn't.

gasstationwithoutpumps said...

SteveH, I think that some of the apparent disagreement in the comments here comes from your non-standard use of the term "test prep". No one here is arguing against teaching a child math, science, or English beyond what the school offers, particularly if the school has rather lax standards.
The term "test prep" is used to mean last-minute cramming for a test and "test-taking strategies", neither of which are as useful as learning the material properly in the first place.

If you need to motivate your son by holding out a promise of higher SAT score in order to get him to study things that are not in the high school courses, that's ok, but it isn't what the rest of us call test prep.

lgm said...

>> I think that speed is huge, not only with calculation of the basics, but in terms of whether you have seen that type of problem before.

My thoughts at the time I took the SAT (pre-recentering) were that speed came from having good enough 3D thinking, good memory, proficiency at multi-step problem solving, mental math proficiency, good logic and good test taking skills. Seeing the specific type of problem before was not necessary to score over 700 but a good prealgebra background was.

Looking over "10 Real SATS" I can see many units that were omitted by my district: sequences, some of the logic problems, some of the prealgebra geometry problems, some of the set theory/operation representations, some of the point on a line problems. All of these will take too much time to work out if the student doesn't have the nomenclature and visual skills.

My 'test prep' all along has consisted of ensuring that the needed content was either taught or figured out, the ability to generate insight was developed, and the visual skills were developed as well as cultivating a liking for solving puzzles. I view the latter as a human trait which bad teaching/environment suppresses.

For ex., taking a cylinder and unrolling it mentally doesn't have to be done before seeing such a problem, but the skill of going from 3D to 2D needs to be there. Factoring a particular expression isn't needed, but skill in factoring and for students who haven't taken Alg I, the ability to extend the concept is. Solving a particular word problem type isn't as necessary as having the skill to visualize and understand the problem plus the underlying math concept.

For ex., from 10 Real SATs....difficulty 5(hardest):
The tip of a blade of an electric fan is 1.5 ft from the axis of rotation. If the fan spins at a full rate of 1,760 revolutions per minute, how many miles will a point at the tip of a blade travel in one hour? (1 mile=5,280 ft).
With all the skills in place, solving this one is obvious without seeing a problem of this type before hand.

I am also interested in hearing other people's experiences.

>>Now I'm trying to figure out how to focus some of his attention away (long term) from his high school classes to preparation for the PSAT and SAT.

Wouldn't the summer be enough?

Crimson Wife said...

I don't consider the term "test prep" to only include last-minute cramming. On the contrary, the folks I know who have been most successful in raising their SAT scores through test prep invested several months in it. And yes, learning test-taking strategies (especially how to avoid falling for the College Board's "traps" when one is unsure of the answer) is a key part of that.

My mom taught Princeton Review so she tutored me on the strategies. I also worked through the Kaplan book (this was in the days before the CD-ROM) and a Barron's book. Plus I worked on memorizing vocabulary (this was when there were still antonyms & analogies) and took all the practice SAT's that the College Board sold.

As mentioned earlier, I raised my final SAT score 200 points from what was predicted by my PSAT.

Allison said...

I'm in GSWOP's camp--"test prep" refers to cramming, Kaplan-like courses, or other classes (perhaps a semester long) designed to concentrate on test taking strategies, practicing of test taking, etc. (I'm ignoring the Verbal SAT test prep. My comments apply to the Math test.)

When you say you're tying to improve your son's test prep over years, I just don't get it why you call it that-- IT'S JUST TEACHING HIM MATH. If he knows math, then the SAT Math test is easy, and there's little need for prepping strategies on how to better guess, and little need to worry about the clock, either.

I'm not clear why my phrase "mathematical maturity" in unhelpful--it would seem to be exactly what you're trying to teach, even though you call it "years long test prep", it seems, though I'd disagree with your notion that you need to look at e.g. The Art of Problem Solving.

I'm much more in line with LGM--success at the SAT comes with speed, and speed comes with success. Speed comes as a result of good memory, proficiency with procedures and mental math, proficiency with multi step problem solving, proficiency at translating each sentence into the proper math statement, and success comes with those things, too. How do you teach these things? By teaching math, until and these things are learned in the process.

Now, the SAT Math does test "up to" algebra 2, meaning there will be a handful of questions on it from those later year high school courses, but the majority of questions will be solvable with a solid prealgebra background. What is gained by taking more advanced math courses, or studying more math on one's own (and maybe you achieve this by afterschooling with TAoPS) is the maturity--making all of that solid prealgebra background so immediate, so unconscious, that there's no need to poll and ask oneself "what question is this like?"

unlike LGM, I can't remember a thing about taking the SAT (in 1987 or 1988), and I'm not very good at visual manipulation of 3d objects, but that never mattered to me. Like LGM, I loved solving puzzles, and enjoyed taking the test. I was extremely good at translating what was written into math, and that didn't require a lot of visual skill if you had other mathematical maturity to handle defns, etc.

VickyS said...

he can focus on learning, not trying to get every last point on his tests

If he's focusing on learning, won't he get those points anyway?

Also, I hate to break it to you, but a lot of colleges recalculate GPA by throwing out all non-core courses. That includes music, so that orchestra honors bump might not mean as much as you hope. They recalulate based on English, Math, History, Science & Foreign Language.

Also...a lot of kids these days end up with GPAs above 4.0, did you know that? I just heard of someone with a 4.7! Our school refuses to go above 4.0 (not that anyone is near that anyway) so my son is already at a disadvantage. But what can you do? These are things we just don't have any control over. Gotta let go.

Maybe just a year to get used to high school is in order here. Your son is just 3 months into 9th grade. If he's enjoying his classes, doing the work and learning the material, even if he gets a couple of B's here and there, is it the end of the world? And he's a smart dude so he probably will get A's anyway!

And I'll state again my contrarian view of the PSAT and SAT, just to give permission to others I guess to buck the system as well. My son did nothing, nada, hardly even looked at the sample questions. I did make him take it a couple of times. He didn't get stellar scores but they were okay (600s for reading/writing, 700s for math). And I didn't write his college essays, either; I didn't even change a punctuation mark. But, he's a good, decent boy, and I firmly believe that these scores and essays are a realistic reflection of the amount of effort and talent he is going to bring to college (at least initially) and he is more likely to attend a college that he can handle. In other words, there will be a good fit, I hope, and my theory is that the fit will be better because we didn't try to position him in any way, not because we did.

Anonymous said...

I went to a high school that had letter grades with no plusses and minuses -- so the grades reported on the transcript were A, B, C, D, or F with no finer gradations. There was no standard system for translating averages to number grades. So in Miss Smith's class a 87 might be an A, but in Miss Greene's class a 92 might be a B. (And this could be fair because Miss Smith might ask harder questions and grade more strictly!)

The difference between doing the minimum to cross over the "A" threshold and doing your best to max out your achievement in the class comes in the recommendations. So someone who coasts after reaching the 93 mark will be described as "more interested in grades than learning" and someone who gives it his all to get a 98 will be described as "always does his best." The teachers talk to each other; they will include impressions about the student's overall work ethic and not just their impression for the one class.

I have been on an admissions committee. We like to see the applications from students who are doing their best and following their passions more than we like the ones who look like they are playing the admissions game. We accepted one kid with spotty grades because his total application was pretty awesome. I have the luxury of having only a small number of applications to read (my states only produce a few hundred applicants), so I have also called up the teachers who have written recommendations. When there are multiple applicants from the same school, sometime I will call up a teacher who has taught all of them -- even if he/she didn't write recs for all of them -- and ask the teacher to compare and contrast. My #1 interest is identifying the high-achieving students who are right for us. We need inquisitive students who are intrinsically motivated by a love of learning.

SteveH said...

"The term "test prep" is used to mean last-minute cramming for a test and "test-taking strategies"

"last-minute"?

By definition? Do you think that anyone at KTM believes that "last-minute" can help more than a minor amount? Isn't it clear that the discussion is much more than that?

Do I just try to teach my son math in some sort of generalized way, or do I base it on a review of several sample SAT tests? As I mentioned, I see questions that my son has never had before. I expect that I will spend some time trying to figure out what the SAT test thinks is important. I'm sure I will be surprised.

SteveH said...

"..how many miles will a point at the tip of a blade travel in one hour?"

OK, but I came across some odd/even problems that I'm sure would slow my son down if he had never seen them before. How about geometry problems and which formulas are more important than others? In some problems, you had to see that a triangle was a right triangle by looking at the lengths of the sides and seeing a 3-4-5 type of pattern. In other right triangle problems, answers were given using radicals, so you better memorize the side relationships of simple angles or you will waste time. These are the sorts of things I'm interested in finding.


"Wouldn't the summer be enough?"

It may end up with that, but I'm trying to figure out a longer-term strategy.

SteveH said...

"If he's focusing on learning, won't he get those points anyway?"

Not necessarily. Getting the most points means trying to figure out what the teacher wants. It means that he should try really hard to get on group projects with kids who will work. If he has a grade over 93, then he doesn't have to worry about that as much, perhaps.


"Also, I hate to break it to you, but a lot of colleges recalculate GPA by throwing out all non-core courses."

I'm quite familiar with that, thanks. However, it matters quite a bit with the class rank and all of the intangibles that colleges will consider.


"Also...a lot of kids these days end up with GPAs above 4.0, did you know that?"

Yup. I'm sure colleges will strip off the weights and map the scores to their own system.


"These are things we just don't have any control over. Gotta let go."

Does it have to be one or the other? To be interested in the details of the calculations means that I will create unnecessary stress and anxiety? There are interesting assumptions being made, and they will be helpful in knowing what can be controlled and what cannot.


"... is it the end of the world?"

Interesting comment. I suppose it might seem that way, but I don't think it's that black and white. I don't believe in last-minute cramming and I don't believe that there is no difference between what is a proper development in math and what is on the SAT. As for high school grades and GPA, my son is quite capable of knowing the details and keeping things in perspective.

gasstationwithoutpumps said...

I posted on this topic in my blog:
http://gasstationwithoutpumps.wordpress.com/2010/11/29/not-prepping-for-sat/

lgm said...

snip ..VickyS>>>If he's enjoying his classes, doing the work and learning the material, even if he gets a couple of B's here and there, is it the end of the world? And he's a smart dude so he probably will get A's anyway!

This strategy wouldn't work here. Earning a B below an 85 means the child will not be allowed to continue on in the math sequence. Most of the time it is NOT the child's fault - it is the instructor and math dept chair who have decided NOT to offer the complete curriculum in hopes of gaming the state test such that ALL children receive a pass, rather than some excelling, some passing, and some failing. The class is slowed down deliberately to get core concepts mastered, and the families that inquire are told that the rest is 'enrichment' and will not be offered this year due to the 'needs of the majority of the class'. If a parent does not offer the missing units on the side, the child will not master the essentials and stay on the college prep trail. End of the world? No, not if he can get the course somewhere else. But why waste the time?

SteveH said...

"So someone who coasts after reaching the 93 mark will be described as "more interested in grades than learning" and someone who gives it his all to get a 98 "

Sigh. Is that what I sound like? That will never happen because it all shows up in the weighted GPA and class rank. Are the advantages to knowing the details overriden by the assumption that the person will use the information incorrectly, or do these details just don't matter?

Overall, a philosophy of doing one's best and not worrying about the details is a good one. However, knowing the details can also prevent unnecessary stress on a daily basis. I would say that the formulas do matter and that there is a lot of low hanging fruit.

lgm said...

RE: 93 vs 98 Many interpretations of this, based on grading scale. In general, 98 means the student has either mastered a lot of the mat'l before taking the course, or has superior study and in-class skills. 93 means improvement needed or cramped for time. "Figuring out what the teacher wants" = in-class skills at gettting the point plus the details. Refer to "What Smart Students Know" by A. Robinson for suggestions.

Re: <>

This should have been done when the course was taken initially. Check out the MathCounts Toolbox for starters on handy math that should be known by heart, then your Brown&Dolciani or Singapore PM & NEM. Essentially the problems posed should be in the 'obvious' category to be considered mastered with fluency.

Anonymous said...

Colleges do recalculate GPAs because there is no common consensus among high schools. And they consider what the rigor of the school and individual schedule are. So, a kid who has all As but hasn't taken a single honors/AP class will not (assuming such courses are available) be looked on more favorably than a kid with As and Bs in all honors/AP courses. Schools can (although not all do) send a profile which puts grades in context.

SteveH said...

In GSWP's link:

"Since my son did very well on the SAT math and critical reading sections when he took them in 6th grade (over 700), I see no reason for him to waste time on test prep for those sections."

Did he take the actual test in 6th grade, or a simulated test? I understand that colleges can ask for all SAT grades from each time it has been taken; that you can't guarantee only the best scores will be submitted. Maybe it's just that if the college asks, you can't deny.

By the way, many parents would see having a child take the SAT in 6th grade as an extreme form of test prep. So, the question is how would you prepare him if he did not get 700? What if his schools don't give him enough practice with 5-paragraph essays? What would you do? Would you work on 5-paragraph essays or a more general approach to writing? Both, I would assume.

Obviously, you had him take the SAT test instead of any old sort of generic test. It seems to me that the test would then drive (part of) the process. I could see a more general approach, however, if the score was much lower.

I think many kids would benefit from a careful "gap" analysis. Many errors could be caused by just a few misunderstandings. My son doesn't have any gaps, but he is not as flexible as he could be. Like most students, he knows the general process to solve a problem because it's in a particular section of the textbook they're working on.

My reaction is that test prep doesn't look good statistically because much of it is generic and done at the last minute. However, long term test prep should not mean just trusting your schools or ignoring what's on the test. To forestall any arguments otherwise, I don't advocate targeting the SAT as the primary tool of education.

My son is a freshman, but he has never taken the SAT or SSAT. I will probably have him take an informal timed practice test at some point this year. I will also go through lots of sample questions to see what kinds of issues he might have. I have the benefit of having taught him algebra and geometry. This again raises the point that while statistics might tell you what is average, they shouldn't tell you what to do about an individual.

You don't have to make the cost/benefit analysis beforehand. You can have a process that includes feedback. One can also discuss different approaches to the task. People can talk about what things made big differences for their kids. These big differences might not show up in the overall statistics, but they can be huge for an individual. Many dismiss anecdotes, but for some, the information can be huge.

SteveH said...

"Check out the MathCounts Toolbox for starters on handy math that should be known by heart, then your Brown&Dolciani or Singapore PM & NEM."

Do you have links? I taught my son geometry, but I was surprised at which formulas the tests expected him to know. I remember one problem where he could have saved a lot of time by remembering the results of a proof.

kcab said...

I agree that talent search SAT/ACT testing is a form of test prep - in fact, that's one of the reasons I had my child take the ACT as last year as a 7th grader. (She'll do SAT this year.) I don't expect she'll do anything other than that to prep though, and she aggressively didn't prep for the early test. (As in, refused to read the instructions book sent to her after registration.) I don't think it's an extreme form of test prep though, it can be somewhat normal, depending on where you live.

Scores from test administrations before 9th grade aren't retained on the permanent record unless requested. Tests taken from 9th grade on do show up on the record.

SBS said...

There are a variety of systems in place regarding which SAT scores a school sees. But, currently, all scores before age 13 are removed evenutally, and normally won't be seen by colleges. If they do show up, the student can just note that they were taken for a talent search.

Crimson Wife said...

The point of talent search testing isn't generally practice for high school testing (though that's a nice fringe benefit) but rather to gain access to distance-learning and summer enrichment courses. When few districts offer any meaningful GATE, programs like EPGY and CTY are an academic life-line for bright students bored by the slow pace and lack of intellectual rigor of heterogeneous courses.

SteveH said...

There are some great comments on GSWP's blog. I liked the one by Josh that discussed different types of problems and the "complete and utter lack of motivation". I saw that in the SSAT class I taught.

Catherine Johnson said...

I need you to come talk to C.

C. gave away **exactly** 3 points in RELIGION class, reducing his B+ to a B (in a class that should be a breeze for him).

He gave away 3 points on the chance to organize his papers into a binder, I think.

Serious talking to here at home following that revelation.

Catherine Johnson said...

The colleges that use rank rather than GPA do so to try to counteract the enormous grade inflation that has happened in high schools.

I'm late to the party, so I don't know whether anyone has brought up the issue of grade DEFLATION in wealthy suburban schools.

See, e.g.: Paul Attewell on Winner-take-all schools.

Catherine Johnson said...

I live in an affluent suburb of Boston. One factor tipping me away from sending my children to a well-regarded prep school like Roxbury Latin is that the same academic performance would result in a lower class rank.

bingo

SteveH said...

In the "Academic Index" (seemingly used for primary sorting of applications), there are three equal parts; SAT-math, SAT-verbal, and class rank, adjusted by the size of the graduating class. Of course, it's very crude to use class size as the correction factor for class rank. This would hurt kids at small, elite prep schools. High schools do, however, send statistics for the school along with transcripts. I would assume that colleges have ways of ranking schools with that information.

What I find more interesting is that the high school grade information is only 1/3 of the overall value of the index. Obiously, the SAT test offers more normalized information than high school grades from unknown schools. However, when you read all of the guidance information at our high school, they say that kids should worry only about working hard in their classes. When I look at SAT questions, I don't see the one-to-one connection at all.


"One factor tipping me away from sending my children to a well-regarded prep school like Roxbury Latin is that the same academic performance would result in a lower class rank."

My guess is that the effect would come later on, when they are looking at the the details of each application. A student who is ranked second at a decent high school will look more interesting (right or wrong) than the 20th clone coming from Phillips-Andover. We decided to avoid this sort of fast-track for our son, not so much for that issue, but to keep him away from the competitive stress. I don't like too much of that kind of motivation.

SteveH said...

I did find out what our high school does for GPA, along with class rank. Actually, what really matters is exactly what gets sent out on the transcripts. They send out a list of classes, the level of the class (gen,CP,hon,AP), the number of credits, and the percentage grade for the class. They also send out the class rank, which is based on multiplying the percent grade by the weight factor of the course and the number of credits. They also calculate an unweighted GPA, but most colleges ignore that if they are given the percent grades per class. So, my advice to my son is to ignore all high school GPA values because colleges will use the other information. Even for class rank, our school doesn't used GPA values. I'm happy with this approach.

My next question to them is whether the school has a grading policy about group grades. Guidance tells students that their class percent average grade is "worth it's weight in gold" and that it should be their major focus. Group grading tells them that it is not in their control.

Another question I asked was about teachers using low grades as a motivating factor, not as a reflection of whether the student applied knowledge and skills that were taught to them. This is the tougher or more vague question. It will be interesting to see what response I get. I will probably get put on an annoying parent list. I don't care. They expect so much from our kids. It should work both ways.

Ben said...

It's tough to know how to best present yourself. When considering GPA, you should calculate it both as weighted and unweighted. Then you can determine which places you in the better light.