Studying the studying study: it’s foreign-language enrollments

About a month ago, the first coverage appeared of a March paper by Philip Babcock and Mindy Marks attempting to pin down changes in college student studying. Since then, lots of people have drawn all sorts of conclusions without reading the paper very carefully. The dominant representation of the paper is the usual declension story, because the top-line conclusion of the authors is that net of all compositional effects, college students are spending less time studying today than several decades ago, in all higher-ed sectors. While the authors spend only two paragraphs on explanations (most of the paper is detailed number-crunching of various surveys, discussions of what’s comparable and what is not), it’s that rank speculation that’s gotten most of the play.

But there’s a substantial problem with trying to fit the top-line empirical conclusion of the study with the standard story of declension, because the standard story of declension heavily weights the supposed decline towards the last few decades. Back when I was a college student, the story goes, we could study Nietzsche while drinking 120-proof Everclear and running 4 miles a day, but the students today want to be catered to, colleges spend all this money on climbing walls and dining-hall furniture, yadda yadda. (For the record, I didn’t read Nietzsche until grad school, I have never had Everclear as far as I’m aware, and I was a very amateur on-and-off-mostly-off intramural badminton guy, not a runner, but I hope you understand my point. See the Monty Python Yorkshireman sketch for the classic example.)

But the paper’s empirical conclusion is that if there’s been a decline in the average number of hours studying (and they claim that decline over 40 years to be about 10 hours per week), that decline occurred largely before 1988. There’s an interesting question here about the question-framing effects of the different surveys (Project Talent in 1961, NLSY’s 1981 follow-up, and HERI in 1988 and 2004), because students were asked to place their studying time in groups of hours rather than state a specific number of hours, and surveys did not consistently ask for the same reference (“typical” versus “last” week). In one way of looking at the data, without trying to adjust for framing effects, the 1961-1981 decline was 4.7 hours, while the 1961-1981 decline was 8 hours after some adjustment for framing effects, but in either case, the last few decades’ decline was minimal.

This substantive conclusion is at odds with most of the representation of the survey. If the study is true, then somehow college students of the past two decades have been less vulnerable to the corrupting influence of pop culture/sports/drinking/frats/whatever than students of the 1960s and 1970s (and possibly 1980s), where the apparent real declines occurred. Or where college administrators in the last few decades have been less influenced by a consumerist orientation, in comparison with the prior generation of administrators. All three of you who believe either story, please raise your hands.

Since Babcock and Marks have made a reasonable prima facia case that the decline happened in every slice they tried to examine (liberal-arts colleges, public universities, community colleges, etc.), and that demographic factors and intensity of study don’t have much evidence as factors, let’s think about the twenty-year period between 1960 and 1980 and the 15-year period between 1990 and 2005, and let’s take their empirical conclusions at face value. What could explain the much larger drop in study time in the first period and a minimal drop in study time in the second period?

• Greater preparation of high-school graduates on average by 1980.
• Greater influence of the “counterculture”: somehow, students of the 1960s and 1970s were more influenced by changing social mores than students of the 1990s and early part of the last decade.
• A loosening of in-loco-parentis attitudes of colleges: Once the ties were loosened in the 1960s and 1970s, there wasn’t much further to go. This is a variant on the prior explanation.
• The spread of electronic typewriters.
• A quantum change in curricular requirements in the first era that would have a much greater change in study time than anything that could have happened in the second era.

The first explanation might work moderately for some elite institutions that could be more selective by 1980, but it’s not plausible for all institutions, especially the growing public institutions in the 1960s and 1970s. The second and third explanations are going to appeal to those who tend to look to cultural explanations, but without changing academic expectations, a cultural argument doesn’t explain how more people could graduate in 1980 with fewer study hours on average than in 1960.

My spouse guessed that the spread of computers might have made studying more efficient. That doesn’t fit with the timing of the drop, but another bit of technology does: electric typewriters. Smith-Corona created the first line of “portable” electric typewriters in the late 1950s, and IBM created the Selectric line (and its little ball o’ letters, so much sexier than the daisy wheel) in the early 1960s. The quotation marks around portable mean that the first Smith-Corona portable was about as portable as the first 25-pound “portable” computers, but you take what you can get at the time. The spread of electric typewriters made it possible for middle-class students to type drafts and final copies without killer finger muscles. (Ha! I know that in the 1960s there were plenty of students whose first drafts were what they turned in.) And without the writing-longhand stage. And with the spread of electric typewriters in K-12 schools, where typing is taught, students would have found learning typing to be less torture than with manual typewriters.

But the electric-typewriter explanation is probably marginal as an explanation. It’s a great Just-So Story, but there’s no way to confirm or disprove it.

On the other hand, at least as I write this entry, I think the most plausible and empirically-based explanation is a specific curriculum change between 1960 and 1980 that you did not see between 1990 and 2005. So where would you look for this? Folks like the late Allan Bloom would look more generally to the humanities, or rather the expansion of professional education at the expense of liberal-arts education and the watering-down of liberal-arts requirements. That’s pretty general, and it doesn’t really explain much about declining study time, unless you can assert that on average professional courses require less work than liberal-arts courses as a generic claim. That might be true but it might not be: ever completed every problem set for a series of accounting classes? No, I haven’t, either, but if the evidence on study time generally is limited, what do you think exists as evidence of different fields?

My speculation on the structural side, and on this you can easily disprove me: changing language requirements that were either reduced or that could be tested out of. There are some types of classes that are very hard to reduce work-time on, at least for most students, and language instruction is one of them. When I attended college in the mid-1980s, my institution (Haverford College) allowed students to test out of the language requirement. So sometime in my first fall, I hopped a bus with a dozen other students, and we rode to an unidentified building at another college to take an acronym-titled test I don’t recall. I passed, and I wasn’t required to read in another language until grad school.

To some extent my speculation is inappropriately based on personal experience, but it jibes with my sense of higher-ed history and my instinct that if there was a quantum drop in study time between 1960 and 1980, there’s a relatively simple and time-specific explanation. And it’s a claim testable through a (somewhat boring) research project: collect the catalog requirements of a range of institutions in 1960 and 1980 and see how curriculum requirements and opt-out provisions changed.

But there is data collected by the Modern Language Association on foreign-language class enrollments, data collected in surveys of foreign-language departments in various years since 1960 (see Table 4, on p. 18 of the 2006 report). The survey year with peak proportionate enrollment was 1965, at roughly 16.5 foreign-language enrollees per 100 students. The survey year with lowest proportionate enrollment was 1980, with 7.3 foreign-language enrollees per 100 students. Since 1980, relative enrollments in foreign-language classes have remained stable or climbed a little bit.

This is an Occam’s Razor explanation of the drop in study time as the result of changing foreign-language enrollments: between 1960 and 1980, average study time dropped because proportionate foreign-language enrollments dropped, and between 1990 and 2005, average study time stayed relatively stable in a time when foreign-language enrollments were stable in a relative sense. No other explanation is needed, whether cultural or economic or what have you. I may be wrong about my speculation on requirements/test-out policies, but I suspect a whole load of people need to apologize for ravings about lazy students and consumer-oriented colleges and universities. (This isn’t a story of great academics from the current generation of students, but it fits with the general historian of education’s assumption of relative stability about human nature and students.) For all their flaws, today’s institutions at least haven’t let foreign-language enrollments erode for a few decades, and enrollments are better than in 1980.

Update: Babcock emailed me this afternoon with the URL of a AEI policy brief he and Marks wrote that’s a reasonable lay-friendly summary of the evidence and an expansion of the speculation on what might have happened. In that brief and in email correspondence with Babcock, it’s clear Babcock and Marks are explicit in stating that some ideas are speculative, such as the student-survey influence on faculty. I’d go further: you’d need some stronger evidence to claim that student surveys were widespread by the mid-70s, enough to cause substantial changes in nationwide behavior of faculty and students.

My suspicion about specific requirements are subsumed in that brief under the “lower standards” catchall category. That could refer to curricular requirements, general classroom treaties resulting in lower expectations, or increasing use of student TAs and adjuncts and the consequences for what really happens in a class. Even within curricular requirements, there’s a question of whether the common requirements are going to influence study-time more than requirements for majors.

Some of that is traceable: curriculum requirements are going to be in catalogs, many of which could be in university/college library special collections (but far more for four-year institutions that maintain institutional archives, so community-college catalogs from the 1960s will probably not have survived in great numbers). Unfortunately, I think the era of public-records statutes and state archives really came in the 1970s, but a few states may have sufficient extant records to show what happened over time across an entire state.