Berkeley vs University of Pennsylvania...can't decide!!!

<p>Does admissions even have the man power, especially at a public school like Cal, to really delve that deeply into every single applicant?</p>

<p>^They don’t, but obviously sakky and mathboy both think they should.</p>

<p>^^ Untrue, and I should make this clear. Not Mathboy. Maybe Sakky. </p>

<p>My view is that Berkeley’s admissions philosophy should be kept the same, but that ideally, it should just admit fewer students than it does. I don’t want a private-school-style admissions process that is unpredictable and wild. I do believe that the UC system is relatively straightforward, and someone with good enough grades and test scores at least can make it into such schools with consistency. They needn’t demonstrate genius at a given academic pursuit, nor need they be star athletes. </p>

<p>It is in private schools that students’ applications are overanalyzed, and offers are given only to those students who fit within the elaborate scheme to produce a certain kind of student body. These schools admit classes, not individuals. UCB appears to admit numbers and statistics, relative to its more discriminating private school counterparts, and I am fine with this. It gives many students in CA a chance at a great school. My point has always been that students below a certain cut where they’re almost certain not to benefit from UCB’s caliber of education should go elsewhere – CSU’s, other UC’s, etc. The point of the public school system is to offer a pretty damn large opportunity for many students to get good education, and I support that.</p>

<p>

</p>

<p>But it wouldn’t be a ‘withdrawal’. Instead, there would be no trace of those courses on your transcript at all. In fact, Berkeley could even deny that you were a registered student at all during those wiped semesters, and reveal the truth only under court order. </p>

<p>Again, I ask, why is it that even a personal bankruptcy can be wiped from your credit report, but bad grades can never be wiped from your academic report? Honestly, what’s worse - doing poorly in class or being financially insolvent? Particularly if the guy isn’t going to graduate from Berkeley anyway, then who cares what his bad Berkeley grades were? </p>

<p>

</p>

<p>Well, frankly, that happens already. Let’s face it. Berkeley students, on average, come from significantly richer families than does the average Californian of the same age. Part of it may have to do with causation - that richer students get more educational opportunities. But part of it surely has to do with correlation with parental culture and genetics: parents who are intellectually oriented will tend to pass the corresponding genes and teach their kids to be also be intellectually oriented, and parents who are intellectually oriented will also tend to enjoy above-average career success. In a similar way, the fact that many children of successful athletes will become athletes themselves has to do with causation: successful athletes can afford to invest in top sports training programs for their children, and they will also pass on the genes that provide for athletic ability and will be able to personally teach their children how to become better athletes. </p>

<p>However, in this case of poorer people like the UCLA girl, the better solution would be for Berkeley to commit to support them with strong financial aid that is sufficient for them to graduate. Either that, or don’t admit them at all. {I understand that the fact that she’s an illegal immigrant makes it difficult to provide her with financial support, as the inevitable questions regarding the proper use of taxpayer dollars would arise.} However, the general principle should be: why admit somebody who you strongly suspect doesn’t actually have the financial resources to make it to graduation? You would just be taking advantage of poor people by having them spend whatever little money they do have and not even get a degree. That’s unconscionable. </p>

<p>Furthermore, the presence of poor students at Berkeley makes my grade-wiping proposal even more salient. After all, poor people - whether due to lack of preparation or simple lack of information about majors - are probably the most likely to flunk out. If anybody would need a fresh start afterwards, it would be them. They should be allowed to transfer to a lower UC or a CalState with a clean slate. </p>

<p>Think about what happens now. If some rich kid comes to Berkeley and gets terrible grades, oh well, at least he’s still rich. He can still probably get a great job simply through his family connections. But a poor kid needs to perform well academically in order to get a good job (or good grad school as a waypoint to a good job) to lift himself out of poverty. Yet right now, some poor students come to Berkeley and not only spend money that they don’t really have, but also have their academic records trashed for the rest of their lives. Hence, Berkeley has clearly made them far worse off than if they had never even gone to Berkeley at all.</p>

<p>

</p>

<p>Computerization is your friend. You don’t need to have individual adcom officers delving deeply into each single applicant. You just develop a computerized statistical scoring system that rates each applicant based on a wide range of characteristics, matched to past student data, to assess whether the applicant is likely to flunk out or not, similar to the same systems that businesses use today to assess and mine giant datasets of customer profiles. For example, health insurance companies don’t have individual agents assessing individual customers in the underwriting process. They utilize statistical software systems to determine, based on your medical history, whether to offer coverage to you at all, and if so, at what price. Similarly, advertising firms mine customer profiles to determine which ad is likely to induce customers to buy which product. Hence, when you and I surf the exact same Internet site, I may be shown one bundle of banner ads, and you may be shown another, as determined by our individual customer profiles. </p>

<p>Nor would such a system be particularly expensive. Berkeley has top-ranked computer science and statistics graduate programs - surely Berkeley could pay for a relatively cheap summer research project for some of those graduate students to develop such a program. {They might even be willing to do it for free if it becomes part of their dissertation and/or they could publish it in an academic journal.} Nor would the costs of the equipment be onerous. Computer processing hardware is dirt-cheap and getting cheaper all the time: just $10k can buy you a cluster that is more powerful than even the supercomputers of just a generation ago. Heck, you probably don’t even need dedicated hardware, as you would just be running the code as a batch program only once a year (during admissions season), and hence you can outsource the processing to one of the cloud providers such as Amazon EC2 or Sun Grid. Nor would the software development costs be large - in fact, they might be free. Berkeley probably already has extra licenses for commercial statistical software packages such as SAS that they’re not completely using (for example, maybe they have 50 seat licenses, but Berkeley is using only 45), and other statistical development packages such as R are available as open-source freeware. </p>

<p>Trust me, this is not that hard to do for somebody who is knowledgeable about software engineering and statistics, as many of the CS & Statistics grad students surely are. Nor would this supplant the adcom. The adcom would retain final power over admissions decisions. However, they would have access to a rating tool to determine, based on past student data, how likely a particular applicant is likely to flunk out. For example, if the software indicates that an applicant is more than 95% likely to flunk out, the adcom should seriously consider not admitting that person for the sake of everyone involved, especially the applicant. But right now, the adcom doesn’t even have that information.</p>

<p>One possible objection I see is that people would be judged on the past information of other people, and some might argue that to be unfair because you’re not being judged on your own characteristics, but rather on characteristics on other people (albeit people who are similar to you), and hence you would effectively be paying for the sins of people in the past. But hey, that’s life. For example, smokers have to pay higher health insurance premiums because past smokers suffered from poor health. Asian women are charged below-average car insurance premiums because past Asian women suffered from few expensive car accidents. Perhaps more saliently, strong high school grades are valued in the college admissions process because past students who earned strong high school grades would also tend to do well in college. Hence, we are already treated according to the behavior of past people, and the statistical mining process would be simply par for the course. </p>

<p>The bottom line is that nobody is proposing that Berkeley become open admissions. No matter what admissions system we utilize, the vast majority of applicants will not be admitted. Therefore the only real question on the table is what sorts of criteria the admissions process should use, and I argue that the likelihood to avoid flunking out should be one of those criteria.</p>