<p>BTW even the UC study finding no correlation between HSGPA and SES stated on page 4 that “Standardized test scores do add a statistically significant increment to the prediction, so that the combination of HSGPA and test scores predicts better than HSGPA alone.” This statement is incredible given that the authors of the study wanted UC schools to abandon using the SAT for admissions.</p>
<p>
The study at <a href=“https://research.collegeboard.org/sites/default/files/publications/2012/9/researchreport-2009-1-socioeconomic-status-sat-freshman-gpa-analysis-data.pdf”>https://research.collegeboard.org/sites/default/files/publications/2012/9/researchreport-2009-1-socioeconomic-status-sat-freshman-gpa-analysis-data.pdf</a> includes colleges that use a weighted HS GPA, with extra points for honors and AP courses. Like all other studies I have seen, test scores were more correlated with SES than GPA was correlated with SES. 25 of the 41 colleges in the study actually had a slight negative correlation between SES and HS GPA. However, all 41 of the colleges had a positive correlation between SES and SAT, such that higher SES groups were more likely to have higher test scores. At many colleges, the correlation between SES and SAT scores was quite large, as high as 0.42.</p>
<p>
</p>
<p>Since the SAT has been around since the 1920’s, that begs the question why high schools in America are not not putting the test on the students’ radar.</p>
<p>
</p>
<p>High schools where few students go to four year colleges may not emphasize college preparation, including scheduling the SAT or ACT, during counseling sessions (especially if the counselors are overloaded to the point that they may ignore the potentially college-ready student who does not bring in the more common issues like being in danger of flunking out or otherwise failing to graduate). Parents who did not go to four year college themselves may not be familiar with the college application process enough to remind students to take the SAT or ACT.</p>
<p>Remember that only a minority (although a large minority) of high school graduates eventually graduate with bachelor’s degrees, so it should not be assumed that every high school student is preparing for four year college or expects to go to four year college. The forum bubble of high schools where 99% of graduates go to four year college is nothing close to the overall reality in the US.</p>
<p>Also, many students take only one of the SAT or ACT, based on the default in their region. This is likely why colleges accept them as “equivalent”, even though they are rather different tests.</p>
<p>GMT, it sure was an extreme comment. Any kid who can find a college website will be pointed to the expectations. And folks, don’t forget the increasing numbers of bright kids at under-performing hs who are getting pre-college mentoring. There are so many assumptions about low SES getting so little attention and such poor educaton.</p>
<p>
</p>
<p>Considering that these forums exist to help confused parents and students (who happen to mostly be from high SES backgrounds and have some familiarity with college already) navigate the college search, selection, and admissions process, wouldn’t you think it would be even more confusing for a low SES first generation high school student in a school where counselors are so overburdened with other issues that they give little or no guidance on college search, selection, and admissions?</p>
<p>I think one issue is that 1st gen kids or those from low performing high schools tend to start the process much later than kids from families where the kids and their friends are expected to go to college. Starting the college search in the fall of a student’s senior year puts him or her at a real disadvantage.</p>
<p>@Data10 You did not follow @Sue22 and our discussion of weighting high school courses for better predictive value. My example of weighting referred to different levels of difficulty. Calculus has a higher level of difficulty than Beginning Algebra. Students taking Calculus should be given higher ranking than student taking Algebra assuming they are awarded the same grade. This differs from weighting AP History vs Standard History courses. In this instance just because a course is labeled “AP US History” doesn’t mean it is any more difficult that Standard US History course.</p>
<p>@Sue22 asked about the helpfulness of weighting courses to allow HSGPA to have higher predictive value for college admissions. I commented that the problem was the cost involved to come up with common method to weight HSGPA to reflect difficulty level of courses. I offered that it was cost prohibitive to implement for all high school courses offered by over 50000+ high schools.</p>
<p>This was the fatal flaw of the UC study. In order to get an accurate relationship between HSGPA and SES, the researchers needed to “weight” HSGPA to reflect course difficulty for every course offered by all high schools by common method and then recalculate the weighted HSGPA for the courses taken by each individual student. This endeavor would have been impractical and if tried would have been very costly. So the researchers did the cheap and easy method of ignoring this vital issue and just correlated HSGPA and SES. </p>
<p>The researchers didn’t bother to at least individualize the data to each school demographic type but took all the data as a whole. This was why the UC study found nearly a ZERO correlation between HSGPA and SES while the @2018RiceParent study that used data from a individual high schools provided a higher level of correlation.</p>
<p>If the courses were weighted and HSGPA recalculated, I believe the result would show that HSGPA and SES would produce a higher correlation than SAT and SES. </p>
<p>One way that researchers could show the effect of SES on HSGPA at least indirectly is to examine the ratio of high income students in gift education or high achiever programs vs those low income students in the same program. If it is anything like my son’s school district, the ratio will fall heavily in favor of high income students over low income students in these programs.</p>
<p>@Data10 Just thought of a way to adjust HSGPA of students of different high schools using existing data. A researcher could examine and compare the percentage of students at a high school who took Calculus. The lower the percentage the greater the deduction from the average HSGPA. </p>
<p>
In the vast majority of high schools, AP classes are more difficult than standard classes. This relates to why such classes generally receive higher weighting in weighted GPA calculations, such as the ones referred to in the study. The more difficult classes are given extra weighting to reflect that difficulty. In many HSs, AP classes can also correspond to completing a higher overall course level upon graduation. For example, the HS I attended only offered AP calculus without an alternative non-AP version of calculus. The highest math level achieved by students who did not take AP calculus was usually pre-calc.</p>
<p>
The previously referenced study mentions weighted GPAs used by colleges, so it likely includes recalculation with weighting done by the college. Maybe the recalculation weighting wasn’t exactly the method you would have used, but it still can be reflect how difficulty weighting influences the GPA-SES correlation. The study found a similar degree of correlation with SES/income and grades between colleges that use a weighted form of GPA with extra points for more difficult classes than colleges than use a non-weighted form of GPA. In recent posts we’ve seen 3 different studies that found a higher correlation between SES/income and test scores than SES/income and grades. It wasn’t just a little higher. In all 3 studies, the SES-SAT correlation was far higher than the SES-GPA correlation. It is higher when you apply extra weight for more difficult courses, and it is higher when you use unweighted measures. It is higher when you look at correlations within a specific HS/college, and it is higher when you look at correlations over a large group. It is higher when you look at correlations among specific race and genders, and it is higher when you look at correlations across all students. </p>
<p>Everything we’ve seen points to a higher correlation between SES/income and test scores than SES/income and grades, yet you’ve repeatedly said you think you’d see the opposite result in various situations. Do you have any evidence at all to support these feelings, such as any kid of research that found a higher correlation between SES and any kind of GPA than SES and test scores, under any unique combination of conditions?</p>
<p>
</p>
<p>Often the official position on course rigor is not consistent with reality. In my area, no senior would believe his compulsory English Lit course is every bit as “rigorous” as calculus, or his history course is as challenging as physics; GPA and LORs can be easily “massaged” by anyone with an IQ in three digits.</p>
<p>100-150 points is a lot, considering the low ceiling in the M+CW portions of the SAT. </p>
<p>So I am back to my two basic questions:</p>
<p>Do the non-submitters earn degrees in math/physics at a rate consistent with the submitters?</p>
<p>Do the non-submitters switch majors to protect their GPA at a rate comparable to the submitters?</p>
<p>Btw, the Epstein paper is excellent.</p>
<p>Not all the colleges recalculate or weight on the front end.<br>
There is a study on soph college gpa related to higher math courses taken in hs. You can find it. </p>
<p>And some of you realize you are forcing a hierarchical system (grades and presumptions about them) into an environment (college) where that measure isn’t standardized? Not across depts, not across the range of colleges. If there were standardized college exit exams, maybe. But there aren’t.</p>
<p>The studies on test optional show minuscule differences in later college gpa. You can seek all sorts of dressing, but other factors were shown to be more predictive of “fit and thrive.” And by folks a heck of a lot closer to the details than CC posters. And who says college gpa is the ultimate factor in assessing education? </p>
<p>And, whether hs kids find one class harder than another is also subject to gross variation. Whether kids hit the college buffet and find a different major than their limited hs courses led them to, whether (yet another) Duke study showed this or that, etc, takes a thread OT. And whether a 60% admit school’s context and explanations somehow suggest universals? C’mon. </p>
<p>But, carry on. </p>
<p>Class of 79 and Leyland,</p>
<p>Went straight here to comment on your posts. Never commented before on any blog ever…but
My husband and I both have Ph.D.s in scientific areas from prestigious U. My husband is Asian and I am not, but I probably have the more “Asian” attitude towards academia. We adopted two girls from China. I feel very strongly about the Asian stereotype, as it has so affected our daughters negatively. I strongly disagree that anyone who works hard can do well on the SAT. I disagree that all it takes is hard work. I quit my very demanding job in pharmaceutical industry to help my girls succeed when it was apparent that they would need me. I spent hours and I mean hours everyday reteaching concepts, making up math problems, reading along with them not because I am a helicopter parent, but because they both needed that much effort in order to become mediocre students. We gave up family life, vacations, friends, careers, weekends, extra activities to get them through school. They could not have worked any harder than they did. The younger one has put up such a rejection to this Asian mentality, that she is now unmotivated to learn, has purposely done bad on standardized tests and is pushing back on us. I think I have done more damage than good with this “kitchen table” mentality. I would say the Asians in this country are on the right end of the bell curve because of genes; the parents had high enough IQs to be in the US in the first place and passed on those genes. It is a very skewed population. Asians are not superior! Hard work then allows those children to reach their full biological potential. I am quite tired of the omnipotent attitude we have been subjected to by Asians. I won’t even talk to one sister-in-law anymore because of her judgment. My younger daughter attends a pressure cooker private IVY feeder school. She has been told by Asian friends that they cannot be friends because the mother told her daughter that my daughter does not have straight As. I have been told by her Asian playmates’ parents that “maybe they can play next summer,” when they have a free moment to spare. Her Asian friends have told her of brutal punishments for bringing home A-s, such as being locked out of the house in the rain. We live close to school and she would bring the girls over for an hour after school and then I would take them back to the late bus. I cannot tell you how many times, the Asian friends called mom and lied, “I am going to extra help and taking the late bus,” just so they could play with my daughter for ONE hour every once in awhile after school. I feel sorry for my girls who have to stack up against the Asian genetic superiority in this country. As a result, my younger daughter has rejected her Chinese heritage completely. They both have been greatly impacted by months in an orphanage which rewired their brains. Good job China. Then they were brought to the East coast into a high achieving family, into a neighborhood where all the kids attend Harvard, MIT, U of MI, Brandeis, Muhlenberg, you get the idea.<br>
If I could have their baby bodies back again, I would do the kitchen table practice a lot less and go outside, teach them how to garden and trouble shoot household problems, travel, take them out of bad schools. They would never know what CTY is, because what is potentially going to happen to all you Asian moms is a lot of resentment toward you later in life. I’m sure you have already experienced this resentment. Believe me, a lot more of it is lurking under the surface. We have one Asian friend whose daughter went all the way through medical school, practiced for two years, and then ran off with a man, got married, stayed home and had babies. Hmm…all that effort was worth it. Good luck with that and I hope their engineering degree is more important to you than the family. </p>
<p>
</p>
<p>Actually, that UC study specifically mentions using unweighted HSGPA over weighted HSGPA because the unweighted HSGPA was a better predictor of college GPA than weighted HSGPA. Obviously, you do not agree.</p>
<p>@adhdboarding
Maybe you’re seeing the wrong type of Asians.
As an Asian (Indian) myself with lots of Asian friends, almost none of them have those sort of traumatic experiences. Sure my parents stress education a lot, but I can assure that even when I got a D one quarter this year (as a junior in high school nonetheless), I was not locked out of the house in the rain and don’t have to lie to my parents to go have fun with friends. I think the vast majority of my friends would agree, although there are probably some parents like the ones you speak of. My mom did establish “kitchen-table” teaching when she was small but hasn’t helped me with a homework assignment much (other than reading 1-2 essays) since middle school. I also find your assumption that it’s because of genes and only smart Asians can make it to the U.S to be both offensive and utterly uneducated about biological processes at the same time. “Smartness” isn’t really controlled that much by genetics and it’s likes extremely complex in it’s genetic details, notice how Albert enstein’s kids aren’t geniuses like him. Also, at least from my experience in india, it’s generally the richer (not the harder-working) folks (think mostly upper middle class) who can get their children to come to the u.s although there are a few struggle stories as well.</p>
<p>@Data10 Since there isn’t a study of the nature that I suggested with properly weighted HSGA, one can look to data on High School Dropout Rates (HSDR) and SES. We would agree that the HSDR is highest in schools located in low SES areas versus high SES areas. We can also agree that generally HSDR is inversely correlated to HSGPA. Higher the HSGPA lower the HSDR. Lower the HSGPA the higher the HSDR. </p>
<p>Thus low SES is positively correlated to low HSGPA and high SES is positively correlated to high HSGPA.</p>
<p>This is apparent in our school district. We have two high schools. One is in an affluent neighborhood with average incomes over $100K and the other in a low income area with average incomes under $30K. The low SES high school has a HSDR that is over 5 times higher than that of our high SES high school.</p>
<p>In our high SES high school about 10% receive free or reduced lunch. There are only 3 low income students in the top 100 of our graduating class of 500 students. If there was no correlation like the UC study suggests then there should have been 10 low income students in the top 100.</p>
<p>Here is an excellent article debunking the value of SAT to low income students.
<a href=“What do SAT and IQ tests measure? General intelligence predicts school and life success.”>http://www.slate.com/articles/health_and_science/science/2014/04/what_do_sat_and_iq_tests_measure_general_intelligence_predicts_school_and.html</a></p>
<p>@adhdboarding,
If the school and its students are half as unpleasant as you make them out to be, then why do you make your daughter go there? Seems like you too have drunk the Tiger Parenting koolaid.</p>
<p>@ucbalumnus Please read prior posts of the problem with weighted HSGPA. The UC researchers found that the weighted HSGPA that was given added bonus because the course was an “AP” or “Honors” was not a predictive as unweighted HSGPA. @Sue22 and I discussed the difficulty of providing true weight for each high school course so a course having the “AP” or “Honors” label may or may not be eligible for the bonus, because it would depend upon the true difficulty of a course. “AP English” may be no more difficult than “Standard English” so no weight would be given to “AP English” just because it had “AP” designation.</p>
<p>The colleges that weighted HSGPA did so based upon whether the course was given AP or Honors and not based upon individual evaluation of course difficulty.</p>
<p>@ucbalumnus Weren’t you the one who stated that in one California school district there appeared to be grade inflation because the modal AP score of an “A” students in the actual AP Test was a score of 1. Such factors in my weighting system would have been taken into account whereas the Colleges that weighted HSGPA blindly gave bonus points because a course had an “AP” designation.</p>
<p>
The earlier claim I disputed wasn’t that HS GPA had a positive correlation with SES. Instead I disputed your claim that grades were more correlated with SES than test scores, as all studies I have seen show the opposite, under all combinations of measurement conditions. This includes ones funded by groups trying to downplay the SES-test score correlation.<br>
The UC study did not find no correlation between SES and GPA. Instead it found a positive correlation, but much smaller than the correlation between SES and test scores. </p>