How do top scorers on tests fail to gain admission to top schools?

tokenadult · August 9, 2007, 10:58am

I noticed that there are various discussion threads here on College Confidential that mention college student score ranges. The National Association of College Admission Counselors (NACAC),

<a href=“http://www.nacacnet.org/MemberPortal/[/url]”>http://www.nacacnet.org/MemberPortal/</a>

the federal Department of Education Integrated Postsecondary Education Data System (IPEDS),

<a href=“http://nces.ed.gov/ipeds/[/url]”>http://nces.ed.gov/ipeds/</a>

and the Common Data Set Initiative

<a href=“http://www.commondataset.org/[/url]”>http://www.commondataset.org/</a>

have collaborated to set common standards for colleges gathering data about admission characteristics of their applicants and reporting data about their enrolled classes each year.

I happened to write about this in a July 8th, 2007 email to two private email lists that include readers of these CC forums. As I wrote then, "By National Association of College Admission Counselors (NACAC) Statement of Principles of Good Practice,

<a href=“http://www.nacacnet.org/NR/rdonlyres/9A4F9961-8991-455D-89B4-AE3B9AF2EFE8/0/SPGP.pdf[/url]”>http://www.nacacnet.org/NR/rdonlyres/9A4F9961-8991-455D-89B4-AE3B9AF2EFE8/0/SPGP.pdf</a>

and by the actual practice of the Common Data Set, colleges report only interquartile ranges for each section of the SAT." The NACAC principle reads like this:

</li>
</ol>

The Common Data Set Intiative Instructions read:

However, as I noted in my July email, "Both the Education Trust college profiles

<a href=“http://www.■■■■■■■■■■■■■■■■■■/default.htm[/url]”>http://www.■■■■■■■■■■■■■■■■■■/default.htm</a>

and the U.S. News profiles

<a href=“http://colleges.usnews.rankingsandreviews.com/usnews/edu/college/rankings/rankindex_brief.php[/url]”>http://colleges.usnews.rankingsandreviews.com/usnews/edu/college/rankings/rankindex_brief.php</a>

suffer from a common methodological error: college median SAT scores are reported by summing the ASSUMED median verbal score (actually the standard score halfway between the 25th percentile and 75th percentile score) and the ASSUMED median math score (same wrong definition of ‘median’)."

Can you all see at once why summing the 75th percentile critical reading score for the whole entering class with the 75th percentile math score of that same group may (probably does) overstate the 75th score level for the (unreported) composite scores of that group? In case this doesn’t go without saying, I’ll post here what I wrote back in July: “But most test-takers have a strong area, either critical reading or math, and thus summing scores from the two sections considered individually probably OVERstates the combined scores of most students at most colleges. At least this error is systematic across all colleges, so that their rank order based on these figures is largely unaffected.” The way to be 100 percent sure, of course, what the composite score distribution is in a particular college entering class would be to report it, based on the actual figures received by the college admission office, but NACAC discourages that. Some colleges report median composite scores anyway, in disregard of NACAC principles. Those reported medians should NOT be assumed to be comparable to calculated medians of composite scores derived by the U.S. News or Education Trust methodology.

It’s sufficient, of course, to look at the interquartile ranges to see if a college has room in its enrolled class for a few more peak-scoring applicants. And once a student wraps his or her mind around how to read interquartile ranges reported for each test section, it is really much more helpful for the student’s planning to know those ranges than only to know a (possibly incorrect) median composite score for a college to which the student may apply.

marite · August 9, 2007, 11:21am

Thanks, calmom. Would the 'disadvantaged educational environment" cover attending Compton High as opposed to Palo Alto High, or does it refer to something else?

Tokenadult: Yes, composite scores are not very helpful (and that’s before considering the utility of the scores).

epiphany · August 9, 2007, 11:40am

“Disadvantaged educational environment” = one or more of:
(1) limited course offerings, materials, teacher certification/preparation, quality of educational facility (modern labs, computers, operative technology within school, etc.)
(2) peer student achievement/expectations
(3) parental educational level & home enrichment, or lack thereof
(4) community of residence: level of culture, libraries, etc.
(5) previous preparation for high school

epiphany · August 9, 2007, 12:08pm

You guys can continue your conversation :), and skip over this. I just wanted to address the comments from mammall about e.c.'s.

I’ll try to situate their importance. (And they are, again, relative in importance, not any more a stand-alone than scores or grades are, nor would they dwarf scores+grades unless the e.c. is so spectacular that it is world-class; if so, you and I probably don’t know those people, because chances are they are people like Yo-Yo Ma and Natalie Portman and Sarah Hughes – the latter whom I believe also had a very fine academic record prior to entering Yale, was it?)

Relevance of pre-college e.c.s to college performance:

(a) independent drive, persistence, commitment
(b) evidence of interest in taking risks
(c) achievement in areas (athletic, musical, visual) which condition the body & spirit, develop the discipline which is essential to independent effort, & which facilitate cognition

Different, more expansive, more encompassing abilities are needed to thrive (not just get by) in college, than is true of high school. Test scores are one measure of that preparation for a different level, but are neither sufficient by themselves to show that, nor exclusive as a measurement of future success.

Areas (a) through (c) above have direct implications for college academics. I didnt make that up. The colleges have discovered that through experience.

Elites Us admit all kinds of highly capable students. Harvard, Yale, and Princeton have each been known to admit students with quite high scores, great grades, and maybe only one long-standing e.c., but particularly gifted in one area, such as math, engineering, creative writing, or studio art (the lopsided student frequently mentioned).
In those cases, the tip was neither the score nor the e.c., (since many applicants have perfect/near pefect scores & several e.c.s) but the exceptional intellectual or artistic gift with tangible indices of that gift.

Also, the term pre-college above is most often more than 4 years. Some students do discover a true, or new, e.c. passion in freshman year of h.s., & rapidly develop it, but more often these e.c.s are more long-standing than that if they are going to carry import as a validating aspect of college readiness. (Not SUBSTITUTING for the academics, but cross-checking the achievement.). Colleges esp. care about the long haul, because that would demonstrate (a) through (c). Activities with longevity carry more weight than some student government position, unless there is add’l evidence of a passion for government that is paralleled in off-campus participation of a significant nature.

An elite college/U is an intellectually demanding academy, but much more than that. It is also a campus with orchestras, choirs, dance groups, debate teams (some of these famous, including international tours), and athletic teams. The ongoing existence of those features are sources of particular interest to alumni and to celebrities in particular fields, who will often contribute to those activities/organizations, or to the U as a whole. Some of the students admitted will have high or perfect scores & contribute to these features. But some of the best contributors to these will not have 2400 scores, yet they will be anywhere from very capable to superbly capable academically, by other measures (GPA matched with the quality of that GPA, & other academic recognitions, on & off the h.s.campus). Further, most applicants and even much of the faculty would want a well-rounded campus experience, to include some of these named features (& more). To eliminate these features (& eliminate some of the imperfect-scoring contributors to these features) would be ALSO to eliminate many fine minds, including some perfect scorers & high scorers who want to enjoy those features as an appreciator or spectator.

An elite college/U is also a laboratory for the discussion of issues OUTSIDE of the classroom. Vigorous participation & interest in campus publications is a critical aspect of such discussions, requiring capable, committed writers/journalists willing to spend time in that in addition to their academics. Activities with concrete attention to such issues (such as community service organizations sponsored by the U) also attract students wth high academic performance, and support the Us desire to show institutional leadership in issues beyond the classroom, but affected by the classrom & discussed in classrooms.

calmom · August 9, 2007, 12:09pm

Marite --“Disadvantaged educational environment” would probably cover going to Compton High as opposed to most California high schools – but it probably would be a good idea for the student to mention something about whatever hardships were faced in the essay.

marite · August 9, 2007, 12:14pm

Thanks for the detailed explanation, epiphany. Also calmom. I think I now get it.

tokenadult · August 9, 2007, 12:21pm

Hear. Hear. That is exactly what colleges have discovered: when they can find an applicant with serious extracurricular involvement of the kind described in epiphany’s post #224, they have a student who will add a lot to the college community as an enrolled student and will do the college proud as an alumnus.

kluge · August 9, 2007, 1:54pm

I agree with Calmom that a student who gets straight A’s in high school and has an 1800 SAT score can probable do just fine at any college in America. I’m also inclined to believe that a B+ student with a 2300 SAT probably can, too. At that level of selectivity I suspect that analysis of comparative collegiate “success” is probably of little utility - which I think is in fact the actual conclusion of the UC study.

I disagree somewhat about the difference between the academic populations of the UC’s and the CSU’s, and the applicability of data from one to the other, particularly as to freshmen, though. CSU’s probably do have a higher percentage of commuters and lower SES students, but that differential is probably least among the incoming freshmen; many CSU students are returning, or community college transfers, etc., and that group probably has the highest percentage of functionally disadvantaged students. The academic distinction Calmom draws is also a lot less than suggested as well. The nine UC campuses have incoming freshman GPA’s ranging from 3.4 (Riverside) to 3.9 (Berkeley), with an overall average of about 3.6 or 3.7. CSU’s systemwide freshman HS GPA is 3.29 - not that far behind. We’re talking about the difference between A- students and B+ students. And there are some CSU’s which draw from the same applicant pool as the UC’s - Cal Poly SLO has a freshman class comparable to those at Davis, Irvine and Santa Barbara; San Diego State’s freshman class HS GPA and SAT averages are comparable to Riverside’s. Those two CSU’s in particular are chosen by many students who were also admitted to UC campuses. So in a sense, for the question of relative predictive strength of grades vs. test scores, I think the CSU data is at least as valuable, and possibly more valuable, than the UC data.

Of course, I think the conclusion to be drawn from both sets of data is the same: test scores seems to correlate more reliably with collegiate success than do HS GPA’s. I wouldn’t have thought that previously. I’m not sure what the significance of that fact is, but I think it’s worth knowing.

FredFredBurger · August 9, 2007, 1:55pm

Depends. What do you define as top schools? HYPSM? That’s too easy…too many qualified applicants. If we are talking about schools outside that realm…it gets harder. Generally these schools have been known to give a nod to the perfect scorer assuming everything else is “in order”. I’d say the reasons would be:

1) Low GPA/Class rank-below top 15%, or 2+ C’s
2) Weak courseload-Not many AP’s
3) Few EC’s
4) And this is the killer-Personality issues

I personally think number 4 is the killer because most kids who score a perfect score or thereabout have high GPA’s, at the very least acceptable GPAs for near HYPSM schools. Also, there’s almost no way you’ll get that kind of score without taking hard classes to train you think critically. Few EC’s can hurt the perfect scorer, but lets be honest, most kids who score that high know what it takes in the admissions process and will make sure to perform meaningful EC’s. I seriously think it’s the arrogant personality that does these kids in. A good portion of them think they are superior to others and/or they don’t help others out (too competitive).

kluge · August 9, 2007, 2:18pm

Tokenadult - you’re right about adding the reported SAT scores for math and verbal to get a average combined score. An easy way to confirm that is to compare the sum of the reported SAT scores to the ACT composite. The ACT score (which is a combined score) is always lower than conversion tables would indicate for the sum of the SAT scores.

I’m not so sure you’re significantly right about using the midpoint between the 25th and 75th percentiles as the “average”. Technically you’re right, but I doubt that the actual average is significantly different from the number thus calculated, and as a casual reference point I suspect it’s “close enough.”

FredFredBurger - I think your dead on with regard to the answer to the OP question - factor 4. I wouldn’t be surprised if such “issues” make themselves evidence in various ways in an application - essays, recommendations, etc.

epiphany · August 9, 2007, 2:52pm

I also agree with post 229, particularly this part, which underscores the falsity of the either/or assumption:

“Few EC’s can hurt the perfect scorer, but lets be honest, most kids who score that high know what it takes in the admissions process and will make sure to perform meaningful EC’s.”

None of the students admitted to true Elites from my D’s demanding private h.s. had e.c.'s as their primary asset, overriding the essential, academic components. Few are admitted every year to such schools. Each of those admits is academically exceptional AND accomplished in e.c.'s – often several, often with longevity, often with high-level recognition/awards in those.

I also fully agree with Fred’s last point about the arrogance factor.

tokenadult · August 9, 2007, 3:09pm

Yes, although the actual median of a data distribution is something to look for by looking at the data themselves, rather than by doing arithmetic with the 75th and 25th percentiles, my comment is more a statement that the people who do the arithmetic should just look at the data directly than a statement that the arithmetic always leads to a significantly wrong answer. Because NACAC discourages the release of median score levels (although college admission offices surely know those for their own college), the technically incorrect but perhaps accidentally factually correct procedure is used by onlookers. So those calculated “medians” may not be SIGNIFICANTLY wrong, but they should be taken with a grain of salt. Many governmental organizations follow the “close enough for government work” standard of accuracy.

tokenadult · August 9, 2007, 3:28pm

I was intentionally vague in my definition of “top” school, but what I mean by that practically in this thread is a highly desired school with an admission profile including lots of high-scoring students. I repeat myself now that this thread has more than 230 posts, but even HYPSM have PLENTY of room to admit all perfect scorers on admission tests. See post #112

<a href=“http://talk.collegeconfidential.com/showpost.php?p=4532526&postcount=112[/url]”>http://talk.collegeconfidential.com/showpost.php?p=4532526&postcount=112</a>

for the numbers.

This comment from your post #229 will be very helpful to other students who read it. I’ll go through each of your points.

1) I actually knew several learners in my school district in my generation who were very smart but not at all expedient about preparing for competitive colleges in high school. They were turned off by boring high school classes (the only kind offered in my school district, alas) and read a lot independently but had horrible grades. Two of the top five SAT scorers in my school district, with three large high schools, dropped out rather than gain their diplomas. Some of the other high scorers, including the very highest scorer, graduated in the bottom half of their high school classes. One of the dropouts actually was admitted to a CC Top 25 LAC, but he eventually completed his degree at the state university, which is where most of the rest of those students attended. Today I guess every high school student knows that the expedient thing to do is to hold your nose, do the busywork, and get good grades, but even today some bright students are unwilling to do that. The most selective colleges still expect successful applicants to have good high school grades.

2) My high school in the 1970s simply didn’t have a strong courseload, not even as strong as the other high school I attended in another state for part of tenth grade. Today it offers much more rigorous classes (which should help with point 1 for bright students who might otherwise be bored) and the issue today in most high schools is to be willing to take the risk of a challenging course, even if that might mean getting a grade of B or two. Something like one out of every twenty high school students take dual-enrollment college classes while in high school

<a href=“http://nces.ed.gov/pubsearch/pubsinfo.asp?pubid=2005008[/url]”>Dual Enrollment of High School Students at Postsecondary Institutions: 2002-03;

so these days a college admission committee would not expect a successful applicant for admission to a top college to have the kind of high school transcript I had.

3) I might rephrase your “few ECs” to “shallow involvement in ECs.” Each year here on CC I see more students who self-report “lots of ECs” get rejected from all of their favorite colleges than students who report DEEP involvement in just two or three ECs. I think what epiphany said above (post #224) strikes the right balance: the idea is to develop oneself through deep involvement in an activity until the student gains abilities the student didn’t have before. Phoning in involvement in twenty different casual student clubs is not the kind of high school experience most highly selective colleges are looking for.

4) Personality issues are a funny thing to write about here, because it is easier to see the mote in the other guy’s eye than the log in your own eye. I agree that some students posting here on CC, and others I know in person, come across as so arrogant that no level of SAT or ACT scores could overcome that bad impression to gain them admission. More generally, I once saw an online comment about a student (not named in the thread) who had a VERY significant level of involvement in a national EC. But the person describing him said he was seriously weird–so weird that he was rejected by all colleges that interview applicants, even though he was accepted by all colleges he applied to that don’t interview applicants. I thought that was interesting background to the varying practices colleges have about offering interviews or not.

Thanks for your post. It will help other students.

You are applying yourself pretty soon–is it this year? Good luck.

collegealum314 · August 9, 2007, 3:53pm

^^I dunno. It depends how you define top colleges as Fred said, but I don’t think arrogance really weeds people out unless you get a rec from someone you rub the wrong way.

I knew a guy whose PhD thesis (in cancer) was so important that it was featured in the national news recently. In high school all he had was outstanding recs and perfect grades at the top school in the state, a team national championship in ARML (2nd most important math contest, teams are by region and you need to qualify to get on your region’s team), 130+/150 score on ASHME (now AMC), very high ranking in the regional math league, German national honor society, and co-captain of the science team that was 2nd in the state. Also did research. Lots of math and science activities, although he was outstanding in every subject. I’m sure he did community service since our school required it and was on a sports team (but didn’t win anything,) but he didn’t really play the college admissions game. No foundations, etc…

There was nothing wrong with his personality, but he was rejected from Stanford and Princeton (waitlisted and rejected on the last one.) I think he definitely would have gotten into some of the other ivies had he applied to more, but still this guy was as close to a sure thing in science as I’ve ever seen.

Had I been making the decisions, I would have taken him over some of the guys who did all the “right” things but were not as smart.

FredFredBurger · August 9, 2007, 4:11pm

I am applying this year. Thanks!

marite · August 9, 2007, 4:20pm

collegealum:

Colleges’ priorities change over time. Stanford accepted my lopsided S with hardly any EC (none that was not related to math/science).

epiphany · August 9, 2007, 4:49pm

"but he didn’t really play the college admissions game. No foundations, etc… "

Nor did any of the admits I’m acquainted with over the last 5 years ‘play [the college admissions] game.’ It’s just that when you start one or more e.c.'s when you are 5, 6, 7 years of age (requesting of parents), and sustain those over another 10-12 yrs., with higher levels of accomplishment, (and because you enjoy both the activity and the improvement), there’s no point in not including those on college apps. Those, combined with their superior academic accomplishments in the sciences, math, humanities, & the arts (with respective national awards in those fields), do make an applicant more attractive. (None of them started foundations.)

Those who were “not as smart” were not chosen over them – whatever e.c.'s they had & whatever “game” those students played, if they played such. But, nice try.

collegealum314 · August 9, 2007, 5:20pm

“**Those who were “not as smart” were not chosen over them **-- whatever e.c.'s they had & whatever “game” those students played, if they played such. But, nice try.”

So you know who got in and who didn’t from my hs enough to make this judgement?

There were very clear cases of “gaming” the system at my school, and sometimes it overcame a lack of intellect. Sometimes it works, sometimes it doesn’t.

bomgeedad · August 9, 2007, 5:40pm

No matter what the theory is, we cannot ignore the fact that the average SAT scores goes up as we go up the elite hierarchy, and within the same university the higher SAT score, the higher is the probability of admission. High score may not guarantee your admission, but it increases your chance.

There are two possible explanation. One is the ADCOM look at scores more than they claim. The other is that there is high correlation between the score and other objective and subjective criteria, so when they look at the other criteria, they end up with students with high scores.

FredFredBurger · August 9, 2007, 5:44pm

^^
Or both explanations.

How do top scorers on tests fail to gain admission to top schools?

CONNECT WITH US