That’s absolute nonsense.
LACs, if you want time (especially, freshman year) to make up your mind what to major in; RUs, if you don’t mind spending freshman year taking a lot of prerequisites. Prerequisites are often an enrollment management (cough, weeding out) tool in disguise.
Do LACs tend to go lightly over content covered in greater depth at typical RUs?
The ACM provides detailed curriculum guidelines for undergraduate CS majors and has assessed coverage of recommended content at various universities and LACs (https://www.acm.org/education/CS2013-final-report.pdf).
It surveys courses at selected schools for hours of coverage in 18 core “knowledge areas”.
For example, for the “AL” (algorithms and complexity) knowledge area, here are the hours of coverage in courses at several universities and LACs:
29-32 hrs in Pomona’s CSC 131
29 hrs in Princeton’s COS 226
28 hrs in Williams College’s CS 256
21 hrs in Grinnel’s “multi-paradigm” 3-course CS introduction
For the “PL” (programming languages) knowledge area:
38 hrs in Pomona’s CSC 131
35 hrs in Brown’s CSCI 1730 and URochester’s CSC 2/454
34 hrs in Grinnel’s “multi-paradigm” 3-course CS introduction
32 hrs in UWashington’s CSE341
31 hrs in Williams College’s CSCI 334
24 hrs in UPenn’s Programming Languages and Techniques I
20 hrs in CMU’s 15-312
For the “OS” (operating systems) knowledge area:
30 hrs in Williams College’s CSCI 432
24 hrs in UArkansas-Little Rock’s CPSC 3380
24 hrs in Embry-Riddle’s CS 420
23 hrs in UHelsinki’s CS 582219
For these and other knowledge areas, it’s hard to make enough apples-to-apples comparisons to support generalizations about coverage/rigor at LACs v. RUs. The few available comparisons suggest to me that in at least some core knowledge areas, some top LACs seem to be offering about the same (or even more) coverage than some RUs. However, it may be the case that some RUs are covering in multiple courses what the LACs cover in just one.
So if you’re willing to pile up on CS courses, no doubt it would be easier to cover these knowledge areas in greater breadth/depth at a larger school. That’s not an approach that most LACs encourage.
This is just one perspective on curriculum coverage/rigor. We also could look at outcomes.
What are the 4y graduation rates for CS majors at LACs v. RUs?
Are students at either one having more trouble getting into all the courses they need to graduate in 4 years?
What about post-graduate employment, or alumni-earned graduate degrees in CS?
I suspect the top LACs do pretty well in these areas.
But, as @ucbalumnus stated in #4 above: each college’s CS department should be investigated individually.
It still comes down to depth and breadth, and this is where LACs (by definition) cannot compete. The LAC - RU in CS debate always comes down to that. Can you get a good, basic CS education? Yes. Is it reasonable to say a comparable student at a RU would get better based simply on the increased depth/breadth that a RU can provide? Again, IMO, yes.
The value of CS majors is not in CS basics. While there are many jobs out there that only need the basics, there are also very many capable programmers out there without college degrees that influence the market demand for “basics” CS majors.
The value and demand is for CS majors with depth in areas where more expertise is required - AI, ML, DL, etc. Acquiring expertise in these areas (even as an undergrad) requires math + CS and access to an in-depth curriculum. I think it’s reasonable to say that an Amherst graduate that has access to only one class in AI will not have the same depth of knowledge as a graduate from an RU that has access to several. It’s probably also reasonable to say that a CS department with only 6 faculty will have difficulty delivering the depth and breadth of a much larger faculty found at RUs.
By accessing the table provided by @ucbalumnus (#18), viewers can see that several LACs offer more key, upper-level courses in CS than a strong university such as WUStL, at least by the standards and timing of that methodology. Generalizations about classes of schools will not apply to all schools in those classes, and might best be avoided.
Not sure if counting hours of coverage is the best way to compare, since some courses may cover more material in an hour than others, due to differences in student ability and preparation. It is really better to go through actual course materials to compare course content if that is desired (though the listings in that ACM document do summarize topics covered in each course). But that can be a laborious task requiring someone with good knowledge of the subject and curriculum.
Anecdotally, Amherst CS majors are often double majors. Since Amherst is unique among LACs in having no distribution requirements at all, there is a lot of room to double or even triple and about 40% of students do complete more than one major.
@ucbalumnus, I’m not sure either that counting hours of coverage is the absolute best possible way to compare (for purposes of a thread like this), but that’s a comparison the Association for Computing Machinery has made available. It’s one perspective. Anyhow, I doubt Williams College students have trouble keeping up and for that reason need more hours to cover Operating Systems than students at other schools do.
One of the themes of the ACM Curriculum Guidelines is that CS content gets packaged and covered in a variety of ways at different colleges. LACs have a somewhat different educational mission than technical institutes or comprehensive research universities do. Their primary mission isn’t highly specialized technical training. If that’s what you’re after, I’d agree, you may want to pick a different kind of school.
Amherst is not unique among LACs in having no distribution requirements. Evergreen State is another LAC that has no distribution requirements (for the BA degree). However, Evergreen State does not really have majors either, and its CS offerings in the areas that one would expect CS majors to learn are extremely limited.
@Rivet2000 Which leads to the question of which LAC’s do offer the best depth and breadth iyo?
What’s your point? Even if there are 2 out of 500 that’s still unique.
And who brought up Evergreen State in this thread? It’s a public college that accepts 99% of applicants and it doesn’t even offer majors, let alone a CS major.
Is it relevant to this discussion in any way?
SMH.
I get it now. When people write about LACs on these forums, they usually mean private highly selective ones. Those which are public or not that selective usually do not count.
“Unique” does mean one of a kind.
Oh for God’s sake @ucbalumnus . You’re the one that needed to make a post “correcting” my comment.
Evergreen State doesn’t have a CS major at all, it doesn’t have single majors nevermind double majors, and no one mentioned it in this thread. It’s not relevant. I can’t imagine why you thought it so important to bring into the discussion.
Regarding LACs and open curricula, the interesting Evergreen State aside, what about Smith, et al?
@Luska19 The only one that I believe would be a match to a student who is all-in for CS is HMU.
Smith requires 64 credits of course work (out of 128 credits for the degree) to be outside of one’s major department. Effectively, it is a large, but very flexible, breadth or general education requirement.
https://www.smith.edu/about-smith/registrar/degree-requirements
So is Smith’s general education requirement a hindrance to having a strong CS program?
Does being “all-in for CS” requires a student to commit more than half his/her college program to the CS major? Some research universities also have general education requirements that amount to ~half the total college program.
I think this is separate from the issue of whether any given LAC offers enough courses or instructors to satisfy the curiosity (or professional training goals) of diverse students. By endowment per student, many of the top 50 colleges are LACs. If they thought it were pedagogically appropriate (and consistent with their missions) to add more courses and instructors, or to decrease their breadth requirements, many of them presumably have the resources to do that. Kenyon’s English department offers about ~100 courses (which would be a lot even for a decent mid-sized university). Granted, the market for CS faculty may be more challenging.
From an Association for Computing Machinery report in 2001:
http://www.acm.org/education/education/education/curric_vols/cc2001.pdf
@tk21769 quotes:
True that. But I think that is reasonable to say that LACs will generally have smaller CS faculty and fewer classes outside the basics. So If you are an aspiring CS student that is seeking in-depth courses in a particular field your best bet is most likely a RU.
By one interpretation, breadth and depth on this thread have in most cases been misapplied. Some universities may indeed offer more CS breadth than LACs, particularly with respect to a few practical programming challenges such as those related to graphics. LACs, however, often excel at the ~10-student special-topics courses that would seem to pertain to CS depth. (I’ll recognize, though, that another definition of depth could relate to multiple courses per advanced sub-topic, perhaps a strength of universities, or, more specifically, tech-focused schools.)
^^Not sure about that. I looked at Amhurst CS classes to examine depth in AI. I could only find one class (offered only in fall 2017). Then I looked for Machine Learning. Again, it lists only one class, and it is not offered in the 2017-2018 year.