139 Possible. There's also a good chance the professsor went to grad school at MIT.
https://qz.com/343521/18-universities-produce-half-of-us-and-canadas-computer-science-professors/
^ āprofesssorā should not type on phone without glasses. :-B
@theloniusmonk , youāre taking one data point and using that to make a generalization. Not a good idea.
What data point, the Jelinek one? I didnāt bring that up, someone else did and I disagreed with the premise, that someone studying under him at JHU gets a PHd there and becomes a professor at a LAC where youāre doing your CS undergrad, is as good as studying CS at a school like JHU itself.
Jelinek being at a top-notch research university and becoming and innovator in NLP proves my point, not refute it. And I understand when Iām making conclusions from anecdotal data.
Most of the arguments in this thread seem to be trying to generalize from one data point (whether it is Jelinek, Stanford, Amherst, etc.).
Fair enough, although youāre slightly misstating my point.
Iām not making strong claims that one kind of school is as good as or better than the other.
Either kind has pros/cons one might want to weigh and consider.
Iām back to the point I made way back in #3:
In general, a LAC will tend to give you more personal attention but a research university will offer more course selection. In general. The poster in #5 apparently thought that statement was āmisinformationā. Iām still not sure why.
But ok, on the RU side, there may indeed be additional advantages such as greater opportunities for direct/indirect exposure to tip-top scholars. The RU benefit may go beyond a classroom connection to specific scholars/professors, if (whether or not they actually teach undergrads, or teach them effectively) their knowledge percolates down through the whole institution. Intellectual trickle-down. The resulting intellectual buzz (if not the classroom instruction) may indeed be stronger at a good research university.
Part of the value proposition for a top LAC is in the extension of teaching to mentored research projects. In the āCompsā process at Carleton, a student writes a comprehensive paper over 3 terms, sits for a 4-hour written exam, or (in CS) completes an ~ambitious project. Hereās an example of a Carleton CS Comps project:
http://cs.carleton.edu/cs_comps/1718/latin/index.php
I think it would be harder at state university scale to manage something like Carletonās Comps process very well. Could a project like that be done better at a place like JHU? Maybe so.
āMost of the arguments in this thread seem to be trying to generalize from one data point (whether it is Jelinek, Stanford, Amherst, etc.).ā
I agree and just for the record, I didnāt bring up any of those points, the flaw in this thread is that just looking at courses is not sufficient when comparing colleges. As the old saying goes, half of what you learn in college is outside of classes. You have a better chance of learning more about CS - how itās used in business to solve problems, how its impacted by hardware, etc. at a tech oriented school (MIT, CMU, Cal Tech) or RU, private or public (Stanford, Michigan, Berkeley, Illinois, Cornell, JHU, Penn) than a LAC.
Love how we keep re-litigating the same points. Sure, if you prefer spending your time around TAs and grad students discussing advances in technology between classes, Stanford, Michigan, Berkeley, Illinois, JHU, and Penn may be your place. If you want one-on-one mentoring from a full-time professor, pursuing their own research, youāre better off at a LAC.
It seems to me that the primary value proposition for liberal arts colleges in general (from numerous threads) is a smaller more intimate learning experience. Unfortunately, that very value proposition limits them in many ways when it comes down to CS in particular.
Smaller faculty means less expertise to draw from (just by sheer numbers), so while LACās should be able to provide great teaching in the basics they will surely be limited in covering the ever growing breadth of CS topics no less depth in those same topics.
Smaller CS classes in general is a plus at any school especially in higher-level classes. However, Iām not sure if there is really any added benefits in the basic classes. While I can see the benefits of a small group of students huddled with a professor sitting in front of a fireplace discussing Kantās Critique of Pure Reason, I somehow donāt see the same benefits if they were instead discussing First Fit vs Buddy System heap allocators.
One-on-one mentoring is great, but to say youāre better off in a LAC is to ignore the fact that more research opportunities are available at larger research schools (hence the tag), and all of the RUs that Iām familiar with promote undergraduate participation.
Cutting-edge CS research happens in greater amounts at RUs (the resources needed are immense), and most ground-breaking discoveries and achievements occur at RUs - and most upperclassmen interested in research are able to get research positions during the school year or summer. In some top RUs it is not even that hard for even freshmen to get research opportunities.
āLove how we keep re-litigating the same points. Sure, if you prefer spending your time around TAs and grad students discussing advances in technology between classes, Stanford, Michigan, Berkeley, Illinois, JHU, and Penn may be your place. If you want one-on-one mentoring from a full-time professor, pursuing their own research, youāre better off at a LAC.ā
Youāve been saying the same thing about LACs as well, so Iāve been very clear on my position - can you clarify yours? Are you saying that there are some LACs that are better than Stanford, CMU, MIT, Berkeley, Michigan, Cal Tech, Princeton, Harvard, Cornell, JHU for CS. If so which ones.
This is what Iāve been saying for the better part of a month now:
ok but personal taste is pretty much sidestepping the question - which is better LAC or tech. Thatās like answering a question of should I eat a banana or an orange, by replying, if you like bananas, eat them, if you like oranges, eat them.
A correct but not very helpful answer. Itās very easy to criticize positions when you havenāt taken any. And you only mentioned two LACs in your reply.
And, youāve only mentioned six universities. When one considers how many more Tier I universities there are in this country (>1000) than there are legitimate LACs (<200) the LACs still come out pretty well. But, we can continue to bait each other ad nauseum. Iām comfortable with what I said nearly a month ago:
Which schools (by type or individually) expect a CS thesis? Though thesis courses have so far been under-regarded on this thread, it would be interesting to see some examples of colleges at which students may commonly complete them.
That is obviously an individual college aspect. Also, there are colleges where a senior or honors thesis is optional.
Carleton, as I mentioned above, has a āCompsā process (for all seniors in all majors, I believe). Each student writes a comprehensive paper over 3 terms, sits for a 4-hour written exam, or (in CS) completes an ambitious project.
At JHU, there doesnāt seem to be a system to ensure every single CS major will initiate and complete a significant project with the dedicated support of a faculty member. However, there is a Senior Honors Thesis Program that looks like it could replicate the Carleton āCompsā experience. It is restricted to students with CS GPAs >= 3.5.
At Princeton, āEach student does at least one design or research project advised by a member of the facultyā (https://www.cs.princeton.edu/ugrad and https://www.cs.princeton.edu/ugrad/independent-work).
IMO, of those three examples, Princetonās model is best. Ideally, every single CS major completes a project/thesis mentored by members of a large, distinguished faculty.
And at Michigan all CS majors complete a Major Design Program or Capstone Course experience.
A thesis/capstone experience is something not unique to LACs or RUs. Most decent computer science programs have such experiences required or easily available.
Hamilton requires a thesis for honors recognition:
https://www.hamilton.edu/academics/departments/Courses-and-Requirements?dept=Computer%20Science
And performed notably well in a recent programming competition against other LACs (Middlebury, Wellesley . . .) as well as larger schools:
https://cs.hamilton.edu/ccscne/
https://www.hamilton.edu/news/story/comp-sci-department-hosts-college-computing-conference
Iām totally not getting some of these arguments. Whatās wrong with saying if you like bananas, eat them, if you like oranges, eat them? Is there a way to definitively show whether a banana or orange is better?