First college application sent

Last night my son got his first set of college applications sent off: University of California, which has its own idiosyncratic deadline and application form. UC does not ask for transcripts and does not want letters of recommendation—students have to enter all their transcript information into web forms.  The lack of letters of recommendation may be a blessing in disguise, as one of his recommenders has still not been able to get the Common App to accept her letter for him. The UC web forms are set up to be fairly easy (though tedious) for students at California high schools, since UC has a list of all UC-approved courses at each high school, but they are really a pain for a home school student.  We were lucky in that his home-schooling was done under a public-school umbrella (Alternative Family Education) that appeared on the drop-down list.  Otherwise, it would have been difficult even to say where he did his high school education.  The instructions for home-schoolers seem to be non-existent and figuring out where to tuck various bits of information was tough.

He ended up applying to 3 of the UCs (UCB, UCSB, and UCSD), though the only campuses he has visited are UCB, UCLA, and UCSC. Why the change?  Well, UCSC is too close to home—he needs to move to more independent living.  Our visit to UCLA made it very clear that undergrads in computer science there got almost no attention from faculty (unless the students were very strong at self-promotion) and acting was mostly restricted to theater arts majors. UCB was better—much better on the acting opportunities, with an attractive acting minor, but undergrads in computer science still had little research opportunity or interaction with the faculty.

We added UCSB primarily because of the College of Creative Studies (CCS) there, an honors college of about 300 students that (the website claims) has close faculty advising and is expected to do graduate-level research as undergrads. The computer science major within CCS looks quite interesting, and (if it lives up to its advertising) may represent a good compromise between the resources of a large university and the attention and nurturing of a small college. Unfortunately, we don’t have an equivalent of the Common Data Set numbers to know how selective CCS is nor does their web site really tell us what they are looking for.

One interesting point is that CCS has a supplementary application that is circulated among the faculty—we regard it as a good sign when the faculty care enough about their program to be involved in choosing who gets in, and when a university allows the faculty to have some say (most UC admissions keep the faculty completely out of freshman admissions—except for coaches at the sports-mad campuses, who seem able to get jocks in even when they don’t come close to being UC-eligible).  Note: transfer admissions at least at UCSC is different, with faculty in the intended department having final say about whether students can be admitted to the major.

UCSD was added as an afterthought, as having a reasonable engineering program while being easier to get into than UCB (38% instead of 17% for male freshmen—UCSB is even higher at 43%).  It is more of a safety school than a careful choice, but the marginal effort of doing an application to it was small—mainly trying to rank the six colleges there based on the very scanty information on the UCSD web site. If he gets in at UCSB or UCSD, but not one of his top three choices, we’ll probably end up doing another visit to southern California, to see how these two campuses feel to him.

The UC applications cost $70 per campus plus another $11.25 each to send SAT scores for a total of $243.75.  He’ll be applying to another 3–7 colleges, so I expect that application fees will end up costing around $1000.  When the cost of college visits and taking the SAT and AP tests in the first place is included, the cost of the application process rises to around $4000–5000.  That seems like a lot, but is dwarfed by the cost of college itself, which for us will be $120,000 to $240,000, depending on which college he goes to—the amount of financial aid that we qualify for seems to vary enormously from school to school.

UPDATE 2013 Dec 1: A reader just pointed out “You can have your official score report sent to one UC campus, and all campuses you apply to will receive it.” http://admission.universityofcalifornia.edu/freshman/requirements/examination-requirement/ I wish I’d noticed that buried in the instructions.  (I’d looked for it, but must have skipped over the line that said it.)

My son, like many high school seniors, has been struggling with the college application essays.  The two he produced for UC seem pretty good to me—one concentrates on the data logger project and is an adaptation of the essay he wrote for the Common Application prompt, while the other talks about why he chose to home school and what that has done for him.  Both essays managed to pack in a lot of information about him and his education, without sounding like laundry lists.

But it took him two weeks to write these essays whose combined length was just shy of the 1000-word limit.  He still has a large number of essays to write (1–3 per college application), and his writer’s block seems to get worse the more important the thing he is writing, so he’s been struggling most with the colleges he cares most about. I have the same problem—I can knock off a blog post like this one in an hour or two, but I have research papers still unfinished that should have been published a decade ago.

The huge amount of time each application takes means that there’s no way that he’ll be applying to the 100s of colleges who send brochures and postcards (most of which are getting recycled unread these days).  Occasionally one of the colleges will send a letter to “the parents of …”, and I sometimes read those for the amusement value, as most of them are so far off target as to be ludicrous.

The main limitation on how many colleges he applies to will probably be how many essays he can get done. I suppose that is why each selective college adds a bunch of essay questions to their application—not so much to find out more about the student as to reduce the flood of applicants to just those who are somewhat serious about attending. This selection process may be counterproductive though, as it would be much easier to churn out acceptable essays for schools he cared nothing about than to try to get a really good essay for a school he cares a lot about.

This weekend, I’m hoping he’ll get the essays done for one of his high-priority colleges (Harvey Mudd or Stanford, for example).

Is college worth the price?

On the homeschool-to-college mailing list (hs2coll@yahoogroups.com) there has been a lot of discussion lately about whether private colleges are worth the $250,000 sticker price for four years.

The discussants generally divide into several camps:

• There is no way that any college is ever worth that much.  Go to a cheaper state college.
• No one pays sticker price—the actual cost is lower at top-ranked private schools than at state schools.
• Why not do two years of community college, then the cheapest bachelor’s degree you can find?
• What you are buying for that huge chunk of money is access to a network of people with money and power—it can be worth it if your career depends on access to such a network.
• When I was a student, …

Of course, everyone is right and everyone is wrong, as is usual in such debates.  The important question is why the student is going to college—what is the point of it?  Only then can you assess whether the price is reasonable or outrageous, and whether it is worth going to college at all.

A lot of people see college as a job ticket—the median earnings of college graduates is much higher than the median earnings of high school students.  If that is the primary goal of a college education, then it is pretty easy to do a calculation to determine the return on investment of buying a college degree.  Lots of people have done that calculation (Google ROI college degree), but I’m not very interested in that view of college, nor with the mindset that sees a college degree as a purchase, rather than as road marker in one’s education.

Another common view of college is that it is a time to make life-long friends (perhaps marriage partners), and that the social opportunities of college are not to be wasted.  While I made some friends in college, and more in grad school, few of them have become lifelong friends—I may see someone from college once every 2 or 3 years, at most. Perhaps if I were a more social individual, that side of college life would have had more importance for me. For my son,  I think that the social aspect of college will be slightly more important than it was for me, but I believe he is mainly interested in finding congenial, talented groups of people with whom he can do things he enjoys (like acting, designing computer programs and hardware, discussing esoteric features of programming languages, learning advanced math and physics, and so forth). [I’d be pleased if he found a girlfriend also, though I didn’t manage that until grad school.] Finding concentrations of congenial, talented people outside colleges and universities is difficult—it’s not that they don’t exist, just that it is hard to find and join them.

Still another view of college is that it is all about book learning—getting a lot of stuff learned in a particular field or in a broad array of fields that mark one as being “educated”. This is a common view for the parents of teenagers in public schools, but a bit less common on the home school lists.  A lot of the home school parents realize that their kids can learn what they want to without needing the structure of a college—the kids have been doing it for years.  Sometimes having a course structure is useful for setting deadlines or determining the scope of a period of study, but other times having a project that needs certain skills and knowledge is sufficient motivation and guidance for the learning.

MOOCs have been a very good deal for home-schoolers, providing the structure of a course without the cost. Of course, the zero cost has a downside—a lot of kids just dabble with the MOOCs and don’t commit to really learning the material, the way they might if they were paying for the instruction.  So even if kids could do the book learning for much less money than going to college, the sunk cost of college tuition provides a powerful incentive to continue and complete the degree.  (Of course most of the sunk cost is coming from the parents, so the parents end up having to motivate the student.)

To determine what college is worth for my son, I’m mainly interested in his long-term well-being—not just the next four years, but the next sixty. I think that he might be happiest if he got a PhD and became an academic, though that career is looking a lot less attractive for the next 30 years than it did 40 years ago, when I was an undergrad—salaries have not kept pace with inflation, workloads have increased, and research grants have gotten much harder to secure.  He might be happy as a working computer engineer or programmer, though I think he’ll find that the workplace is full of only marginally competent people, unless he gets very lucky in where he works.  I think he’s more likely to find congenial workmates if he joins a startup with people he has worked well with for some time, than if he joins an existing corporate structure.  But if he works for a startup when would he have time for acting (or a life)? It would be easier for him to find a group of competent people to do a startup with at some colleges than others (Stanford, for example, has a culture of creating startups, some of which have even been spectacularly successful).  Making a successful startup without colleagues found in college seems much riskier to me.

Most of the careers that would hold his attention require at least a bachelor’s degree, so I’m beginning with the assumption that he’ll do at least four years of college.

The University of California campuses provide a pretty good education for about $32,000 a year (figure $130k–160k for 4 years when tuition inflation is included).  That looks like the cheapest path we’d consider.  The cheapest route (2 years of community college, then transferring to a state school) is not likely to work for him. He has taken some community college courses as a high-school student, and they were fine for him then, but he is unlikely to find many colleagues at his intellectual level in a community college.  If it were the only way he could afford to go to college, then he could make it work, but we’ve been saving 10% of household pre-tax earnings each year since he was born in a 529 college savings plan, so we have enough saved to give him more attractive options.

Based on “net price calculators” from top-notch private schools that my son is considering, we would be expected to pay $42k–61k a year (figure $180k–$250k for 4 years). Incidentally, those net price calculators vary a lot in their ease of use.  Several schools use a net price calculator that can use information stored with the student’s College Board login—those are the easiest to use once you’ve filled one of them out, since the information does not have to be looked up and re-entered for each college.  Unfortunately, many of the colleges my son is considering don’t use that method, so I’ve not done net-price calculations for all of them.  Some of the colleges also include some merit aid in their packages, which means that the net-price calculators are too inaccurate to get more than a very rough idea of price from anyway.

At some schools we’d be expected to pay the full sticker price, while at others the discount is about $10k. Note that some colleges expect us to pay over half our household’s pre-tax earnings each year—they really penalize you for being thrifty and saving for college.  There undoubtedly exist schools that would give him a full merit scholarship, but they are unlikely to have enough intellectual stimulation in the fields that interest him to justify the time, even if the tuition is covered.  None of the schools we looked at provided anywhere near that level of merit aid.

Is the extra $50k–$100k for four years of a private college worth it? 

Well, it depends, of course.

In some colleges, what that money buys is “country club” dorms and activity centers—which don’t interest him very much.  Given that he is not a natural socializer, the network of political connections offered by some colleges is also of little value to him.  What I believe he values is working on projects too big for one person (both in theater and in engineering) and easy access to research opportunities. Some colleges provide more of both project-based learning and undergraduate research opportunities than others, and it isn’t necessarily correlated with public/private or big/small.

The University of California campuses do not provide as much of either projects or undergraduate research as most of the private schools he is looking at.  Although the UCs  have a lot of research, the ones he would apply to also have a lot of grad students doing the research, and the openings for undergrads are not proportional to the huge numbers of students who might want them.  Just finding out about research opportunities, which often requires personal contact with the professor leading the research group, could be difficult. (I think that UCSC does much better at providing research opportunities to undergrads than the other UCs, probably because the grad students only make up 10% of the student body, but the faculty are expected to be just as productive in research as at the other UCs—so faculty have to include undergrads to have enough people working in the lab.)  The classes at UCB and UCLA (the 2 UC campuses we toured) are huge in computer science, with upper-division courses having 50–200 students in them.  It is very hard to make connections with the faculty or with other students in such huge courses.  Acting at UCLA also looked to be almost impossible for non-theater majors—even theater minors couldn’t get roles (according to the theater majors we talked to).  The acting minor at UCB looked much more reasonable, though.  So access to faculty, to projects, and to research may be better at several of the private schools we looked at—that could be worth the price differential.

I also think that some of the private schools are more likely to provide the emotional support that my son might need to get through 4 years of college.  I’m not thinking of formal services like counseling or health care, which may be just as good or better at the larger schools, but of a sense of community and students feeling like family, rather than a random collection of strangers struggling alongside one another.  Different campuses really had a different feel to them when we visited.  Some felt like family, some like business partnerships, some like academic competitions, some like sports bars, and some like ant heaps with 1000s of anonymous, nearly interchangeable individuals. I think I’d be willing to pay more for a college that felt like it would be a home for him, rather than an office or a study cubicle (and certainly more than I’d be willing to pay for him to go to a sports bar).

The bottom line is that we saved enough for him to go to a private college, and if he gets into one that is a good fit, we’ll pay what it takes for him to go.  I think that the extra value is there in a few schools to justify the increased price.  Of course, there are also many, many schools that don’t look like a good fit, and that would be worth less to him than UC, but he is not planning to apply to any of those, despite the pounds of paper they have sent him in postcards and brochures. He’ll also be applying to a few UC campuses, as “safety” schools, since almost all the private schools he is applying to have very low admissions probabilities.

Udacity moving into corporate training

I just read a good article, Udacity’s Sebastian Thrun, Godfather Of Free Online Education, in which the author claims that Udacity is abandoning the MOOC vision of free university-level education in favor of the more lucrative field of corporate training:

It is a good story, as well manicured as a college quad during homecoming weekend. But there’s a problem: The man who started this revolution no longer believes the hype.

I can’t say I’m surprised.  I’ve always regarded the MOOC as more of a PR gesture than an enduring way to provide college education, and I predicted that the companies doing MOOCs would drift into corporate training after burning through their initial corporate capital (if they didn’t simply fold or downsize to the level sustainable as a purely public-relations game).

Long weekend, little accomplished

I just had a 4-day weekend, in which I got little more accomplished than a usual 2-day weekend.

Much of Saturday was spent trying to use PacBio reads to improve a draft genome of  a V. cholerae strain that  I had built with 454 reads a couple of years ago.  There was no problem getting blasr to map the reads, and I could call variants with “samtools mpileup”, though that took 2 CPU days to complete.  Unfortunately, that did not tell me what I really needed to know, which was whether the orignal assembly was in the right order.  I found a couple of places where the PacBio read mapping indicated problems (either the reads all terminated their mapping at nearly the same point, or they suddenly switched from aligning very well to aligning poorly).  Unfortunately, I’ve not yet figured out a good way to automate this detection, so I’m not sure I can find all the places which might have problems.  Dips in average quality of the mpileup consensus over 50- or 100-base windows pulls out the places where the alignments get bad, but not where they suddenly stop.  Furthermore, once I’ve identified the bad regions, I still need to break the genome apart there, rebuild the bad regions from the PacBio reads that map nearby, and see if I can stitch the genome back together (probably with extra repeats that had not been resolved in the 454 assembly).  I’m considering backing off and building a new genome assembly from just the PacBio reads (after cleaning them up using PacBio2CA and the 454 reads) and the Celera Assembler.  I can then compare the genome built from the PacBio reads and the one built from the 454 reads and resolve any discrepancies. Sigh, this project keeps getting bigger, just as I think I’m almost done.

On Sunday, I did a bunch of small tasks: raked leaves and shredded them, updated the grad alumni web page, announced the Freshman Design Seminar class (which will happen winter quarter, though once again as a “Group Tutorial” to prototype the course before submitting the official paperwork), wrote a letter of recommendation for a student applying to grad schools, scanned in the flyer for “Planet of the Abes” (the recent Dinosaur Prom show), scanned 35-year-old t-shirt of mine so that I can get another copy made, updated my paper list to include the just released PNAS paper, wrote a blog post, and caught up with a lot of my e-mail (though there are still some advising e-mails that I haven’t taken care of).

The flyer, drawn by Hunter Wallraff who holds the copyright, for the Dinosaur Prom Improv show. Because the edge of the drawing was not reproduced on the flyer, I had to try to finish the S and add an E  in by hand for “Broadway Playhouse” to get something usable for the titling of the video. I did not correct the error in the URL for westperformingarts.com

On Monday, I did a lot of grading, wrote another blog post, used the Planet of the Abes flyer to make titles for the Dinosaur Prom video, and rendered the video (tying up my laptop all night).  I also cleaned up the scan of the old t-shirt and converted it to SVG so that a new silk screen printing can be done. I’ve tried looking for the copyright holder for the design, but I have no idea how to find him (or her)—Google image searches bring up nothing similar, and there is no signature on the design or the shirt. I started working on my slides for the talk I have to give on Thursday, but did not get much done.

This morning, I responded to more e-mail, wrote another blog post, did more grading, and returned the activity monitor I’ve been wearing for the past 2 weeks to the Sleep Center. In the afternoon, I did more grading at Gayle’s Bakery in Capitola, met with my son’s consultant teacher for a couple of hours, bought the usual weekly load of soy milk (only 2.5 gallons this week), did some other grocery shopping, finished the grading, recorded the grades, cleared the rest of the advising e-mail, and compared results on group theory problems with my son.  We’re a bit behind schedule there—he’s not finished all the Chapter 1 and Chapter 2 problems I assigned, and we’re supposed to be finishing Chapter 3 this week—I’ve not even assigned Chapter 3 problems yet.

Things I wanted to do this weekend but didn’t:

• Get the slides done for Thursday’s talk
• Get the Program Learning Objectives written for bioinformatics
• Get assessment plans defined and written for the Program Learning Objectives for both bioengineering and bioinformatics.
• Create a draft of a revised curriculum for the third track in bioengineering (which also needs a new name and a clearer focus).
• Rewrite the handout for the next programming assignment in the Bioinformatics: Models and Algorithms courss.
• Write code for looking for regions of the Helicobacter pylori genome that are possibly swapped in the current assembly and test for which rearrangement is most consistent with our data.
• Start testing the BitScope differential input device they sent me.
• Start working on Chapter 3 problems in group theory.
• Start writing a paper on the segmenter that I described in my blog 3 months ago.
• Clear the leaves off the roof before the rains start, since the leaves form dams that keep the rain from running off into the gutters properly.

There were probably other things, but I forget what they were now.  Once the to-do list gets longer than my piece of paper can hold, things fall off it.

 

Grading programming assignments

In Critiquing code, I mentioned that I spend a lot of time reading students’ code, particularly the comments, when grading programming assignments.  For the assignment I graded this weekend, I did not even run the code—grading students entirely on the write-up and the source code.

The assignment is one that involves writing a simulation for a couple of generative null models, and gathering statistics to estimate the p-value and E-value of an observed event.  Running their code would not tell me much, other than detecting gross failures, like crashing code.  The errors that occur are subtler ones that involve off-by-one errors, incorrect probability distributions for the codon generator, or not correctly defining the event that they are counting the occurrence of.  These errors can sometimes be detected by looking at the distribution of the results (particularly large shifts in the mean or variance), but not from looking at a small sample.  The students had available plots of results from my simulations, so they could tell whether their simulation was providing similar results.

So I read the write-ups carefully, to see if the students all understand p-value and E-value (this year, sadly, several still seem confused—I’ll have to try again on improving their understanding), to check whether the distributions the students plotted matched the expected results from my simulations and previous years’ students, an to see whether the students explained how they extracted p-values from the simulations (only a a couple of students explained their method—most seem to have run a script that was available to them without even reading what the script did, much less explaining how it worked).

Whenever I saw a discrepancy in the results, I tried looking for the bug in the student code.  In well-written, well-documented code, it generally was fairly easy to find a bug that would explain the discrepancy.  In poorly written, poorly documented code, it was often impossible to figure out what the student was trying to do, much less where the code deviated from the intent. Even when the results appeared to be correct, I looked for subtle errors in the code (like off-by-one errors on length, which would have been too small to appear as a change in the results).

There was only one program so badly written that I gave up trying to figure it out—the student had clearly been taught to do everything with classes, but did not understand the point of classes, so he turned every function into its own class with a __call__ method.  His classes mostly did not have data in the objects, but kept the data in namespaces passed as arguments.  The factoring into classes or functions bore little resemblance to any sensible decomposition of the problem, but had arbitrary and complex interfaces. The result looked like deliberately obfuscated code, though (from previous programs by this person) I think it represented a serious misunderstanding of an objects-first approach to teaching programming, rather than deliberate obfuscation.  I instructed the student to redo the code without using any classes at all—not an approach I usually take with students, but the misuse of classes was so bad that I think that starting over with more fundamental programming concepts is essential.

Some of the students are now getting fairly good at documenting their code—I don’t seem to have any superstar programmers this year, but there are a couple of competent ones who are likely to do good work on their theses (assuming they don’t discard the documentation habits I’m enforcing in this course).  Some of the students who started out as very poor programmers are beginning to understand what it takes to write readable, debuggable code, and their programs have improved substantially. Many of the students have figured out how to separate their I/O, their command line parsing, and their computation into clean separate parts of their code, without sacrificing much efficiency. Even several who were writing very confusing code at the beginning of the course have improved their decomposition of problems, simplified their style, and improved readability of their code enormously.

In one comment on Critiquing code, “Al” commented “I would guess it has more to do with grit and the ability for a student to stick with a tough problem to find a solution.” I rejected that interpretation in my reply: “some of the worst programmers are putting in the most effort—it is just not well-directed effort. The assignments are supposed to be fairly short, and with clean design they can be, but the weaker programmers write much longer and more convoluted code that takes much longer to debug. So ‘grit’ gets them through the assignments, but does not make them into good programmers. Perhaps with much more time and ‘grit’, the really diligent students could throw out their first solutions and re-implement with cleaner designs, but I’ve rarely seen that level of dedication (or students with that much time).”

Now I’m not so sure that my reply was quite right. Some of the biggest improvements seem to be coming from students who are working very hard at understanding what makes a program good—when I complain about the vagueness of their variable descriptions or their docstrings, they improve them in the next assignment, as well as redoing the programs that elicited the feedback. But some of the students are falling behind—neither redoing assignments which they got “redo” on, nor keeping up with the newer assignments.  So there may be some merit to the “grit” theory about who does well—it isn’t predictive for any single assignment, but it may help distinguish those who improve during the course from those who stay at roughly the same level as they entered the course.