Friday, March 14, 2014

Early high touch intellectual experiences for students

One of the peculiarities of any student information system, as an instructor sees the data in it, is the profile information available about each student.  Major and year in school (freshman, sophomore, etc.) might seem like obvious things to include and in my experience they are.  I wish they also included prior courses taken and grades in those courses, as it would give a better picture of the student's preparation as well as provide an indicator of the student's grade expectation.  But that information is FERPA protected, so instructors can't see it.  Academic Advisors can see it.  This sort of difference in access to profile information can't possibly be in the law itself.  Instead, I believe it results from institutional policy, the mindset of which seems to say that instructors don't advise and they are to teach students irrespective of their prior preparation.  It is the student's call whether they are prepared for the class or not.

I am making a thing of this, though I am resigned to the fact that the situation will not change, just to emphasize what I say next.  The student's home address is available to instructors in our SIS.  I have no idea why.  In principle, this allows the instructor to know which students are from the Chicago suburbs, which are from the city itself, which are from downstate, and which are from out of state.  In general, I ignore this information, but for a student who has come to my attention, for good or bad, I will sometimes look at it to see if it correlates with what I have observed.  For example, suppose I have an out-of-class meeting scheduled with a student who then blows it off.  If I find the kid is from a well-to-do suburb that fact helps to complete the picture to explain what's going on.  At an entirely different level, the home address information provides a reasonably good indicator of what the student's family is paying in tuition. 

That different students pay different tuition rates for what seems to be the same service amounts to price discrimination, though one must be careful in thinking about this.  Are there really cost differentials in serving the different students that explains the difference in tuition rates?  For example, the difference in tuition rates between in state and out of state students can be partly explained by the former receiving a subsidy from tax revenues while the latter do not.  The price discrimination piece is only that part of the out of state tuition which can't be explained by the lack of subsidy.  My sense is that such price discrimination has increased over time as part and parcel of the rise in tuition more broadly and the influx of international students at the undergraduate level.  It is one way the university has remained financially afloat as the tax revenue dollars have not kept pace with the rest of university expenditure. 

Price discrimination often comes with quality variation as part of the deal.  One might feel it anathema that quality variation in the education supplied be based on the home address of the student.  Yet there is other type of price discrimination with associated quality variation at the university where most people hardly raise an eyebrow.  One example is differential tuition rates across colleges, with some colleges requiring a tuition surcharge over the base tuition while Liberal Arts and Sciences, in particular, does not.  The colleges with a surcharge typically have better facilities, better ancillary services, a higher expected wage upon graduation, and perhaps other quality enhancements.  They also have higher average standardized test scores of the entering students, which suggests these quality "rewards" are merit based rather than tuition based.  I will return to the issue of merit rewards below.

Perhaps surprisingly to most people, until they think about a bit, price discrimination happens even under apparently uniform pricing, because of quality variation in delivery.  Upperclassman typically are in smaller classes and are more likely taught by tenure track faculty.  Freshmen are more apt to be in large classes, now frequently taught by adjunct instructors, and have graduate students as TAs.  Indeed, when I first started at Illinois there was a tuition differential by class rank.  Seniors paid more than freshman.  At some point they got rid of that differential, the consequence being that we now have flat tuition for all four ranks (though it does vary by the year a student starts) but with quality generally rising as the student gets further along in her studies.  For the same reasons, there is a kind of price discrimination in favor of students from well off suburban high schools and against students from poor rural and inner city schools.  The suburban kids typically get to take a lot of AP classes and if they score well enough on the AP exam then get to bypass that particular large lecture class.  They also are apt to graduate earlier and therefore pay less tuition overall.

Let me turn to non-tuition based variation in educational quality that the institution embraces.  One of these is via the various Honors programs - there are programs at both the campus level and the college level - which are targeted at the very high performers among the students.  So Honors programs fall into the merit reward category.  In the previous decade I taught three times in the Campus Honors Program.  Small seminar classes with very bright and engaged students are a joy to teach, but there is a reason why most of the rest of our course offerings are not taught in a small seminar.  Then, even with the very limited number of CHP offerings, the model was for faculty to teach a CHP class is to do so as an overload and earn a modest stipend from that.  Further, because the size of the program is quite small, there is the question of where the line is drawn for admission to the program and if there are quite meritorious students who don't get in because space is limited.  The last time I taught a CHP class, in fall 2009, I queried the class on this point.  The consensus was that in their regular classes there were some students who did just as well as they did but were not in Honors.  For illustration, though I'm making up these numbers, if CHP includes the top 2% of the class, then perhaps the top 10% of the class is meritorious. Additionally,  there is some measurement error in properly coming up with rankings like these, so who gets in can be a bit arbitrary.

Then there are non-tuition based interventions targeted at disadvantaged students.  I participated in one of these for I-Promise students, as a mentor.   All mentors volunteer, but as far as I know volunteer mentoring is not generally available to students.  In this particular case the idea of the mentoring is to focus on the first year the student is on campus.  If the student can get over the hump of that first year, then the student is apt to get through.  The I-Promise kids are academically talented, but from low income families.  The mentoring is aimed at helping the students negotiate through some issues that most other students won't confront at all or, if they do confront them, then they have other buffers to manage the issues.  That rationalizes the program as is.   But one might reasonably ask whether students from reasonably well off families might benefit from mentoring too, thinking that there are more dimensions to student performance than not dropping out of school.  If so, why don't we try to do something in this arena?

There is another way instructional quality varies from student to student and, indeed, such variation happens within an individual class, at least in my class.  Of course, some of the instruction is pushing stuff at the students and in that sense the effort is uniform.  But another part of the instruction is reciprocation for something that the student initiated.  Students who initiate a lot get more and better reciprocation.   Passive students who initiate hardly or not at all don't get as much of my time.  Such initiation can happen in many different forms.  One is by asking questions during class.  Another is by asking questions online outside of class.  A third is by attending office hours on a regular basis.  A fourth is by making an unusually good contribution in completing course work.  Still another comes from completing assignments early and getting noticed for that.  Maybe sitting in the front row is a form of initiation.  (Surely sitting in the back is the opposite, though my classroom last semester was pretty small.)

If all student behavior were rational, if it were understood that initiation by the student would bring out some reciprocation by the instructor, if that reciprocation were perceived by the students as higher quality of instruction, and if the students wanted higher quality of instruction as long as it didn't impact their pocketbook, then one should expect the total capacity of the instructor to reciprocate to get used up and that the students would compete for that bit of reciprocation which is directed at themselves.

This has not been my recent teaching experience.  Based on what I have observed in the past three years at least one and possibly all of those conditionals don't hold true, as a general matter.  In contrast, a handful of students do act in accord with the model described in the previous paragraph.  They have been socialized into acting responsibly about their own education.  The system works for them.  It doesn't work so well for the others, the majority of students.  What should be done about it?

High touch, face to face interaction with a credible peer, is the answer.   You don't have to take my word on this.  Atul Gawande has a really excellent essay on the diffusion of innovation from last summer called Slow Ideas, which makes this essential point.  His context is different from what we're considering here.  Gawande is concerned with proper infant care by mothers in developing countries.  The mother is unaware of healthful approaches and often does things that are harmful to the baby, but the mother does this out of ignorance, not out of anger. If the mother understood a better approach, she'd embrace it.  One would think the analogous thing should be true for our undergraduates.

In anticipation of Gawande, really in borrowing results from 1990s about technology innovation in high enrollment courses, where to me the most interesting single capability the technology enabled was to put undergrads who had taken the course before in a role of responsibility in assisting students currently taking the course, I wrote a series of blog posts back in 2005 entitled Inward Looking Service Learning (there are 7 posts and if all of them read they should be read chronologically, though they are displayed in reverse chronological order at the link) which sketched how a solution scalable to our campus might be attained.  I still subscribe to the thesis, though the particular mechanism in those posts, the focus was on the study group and peer-mentors would interact with current students in that setting, might not be as effective as one-on-one interaction.  However, one-on-one interaction would be much harder to manage at scale.  In other words, there needs to be a lot of experimenting on the right way to do this.

Nearly nine years after those INSL posts, however, nobody seems to be embracing the idea.  There are a handful of dedicated instructors who subscribe to peer mentoring in their own teaching, finding the approach via their own research, not though my blog.  But the peer mentoring idea is not diffusing at all beyond that, slowly or otherwise.  And in its current usage peer mentoring is about teaching a particular class better.  It isn't (yet) about socializing students to become more active about their own learning in all the classes they take.   Higher ups and people in the learning technology field alike don't seem to be thinking of the issue this way. 

The suggestions below are meant as ways to raise awareness of what early high touch approaches might accomplish.  One of the issues here is that doing anything novel that otherwise doesn't reduce obligations from elsewhere can only be done on a volunteer basis, or possibly on an overload basis if some non-recurrent dollars can be found to support experiments in this area.

  1. Find a group of willing instructors who teach high enrollment courses that are taken primarily by first-year students and do such teaching on a regular basis.  The disciplines of these courses can vary.  Have these instructors form a discussion group based on the following.  Each instructor will have a high touch discussion section that they will run - say with 5 students or 10 students at most.  Have the students opt into the high touch section, with the understanding ahead of time that they will have some intensive interaction with the course instructor.  Let the experiences in teaching these sections be the basis of the discussions the instructors have with one another.  Because of the selection bias from how students opt into these sections, avoid trying to compare performance of the students in the high touch section with the rest of the students in the course.  The focus should be on whether the students in the high touch section seem to grow over the semester and also on whether the instructors get some insight into their own teaching as a consequence of these interactions.  
  2. Do something similar with Teaching Assistants.  In addition to their other sections, let them have one that is high touch (here I'd suggest no more than 5 students in that section).  Have each TA in some course do this and then have the TAs interact about this experiences in the high touch sections in a similar manner to how the instructors interact in the discussion group. 
  3. In both of the above this would be new and the fear is that the participants would revert to the old way of doing things - they'd lecture to the few students who are there or if not that then totally dominate the conversation when in one-on-one discussion with students.  This can be anticipated.  I am not sure what prior preparation would best discourage it, but perhaps some modeling of peer mentoring by current practitioners should be made available to the participants ahead of time so they have some mental model ahead of time of what effective practice looks like.  
  4. There needs to be several neutral observers who can both write about what happens in these experiments and communicate to the powers that be about whether it is worthwhile to do (fund) additional rounds of experimentation.  These observers should also engage in focus groups with the students.  The students, perhaps more than the instructors, might be better able to suggest changes in practice that would encourage them to improve their learning strategies, this in spite of the selection bias mentioned above, because the students will have perspective about the consequences of taking other courses where the high touch approach is not embraced.  Anyone who confronts the scaling of such a high touch approach will have to reckon with the fact that students who participate early on will mainly be operating in the traditional environment and that might overwhelm high touch interventions which would be effective if they were more prevalent.

Let me close with some acknowledgements.  I had lunch with Joe Hinchliffe on Tuesday and with Burks Oakley on Wednesday and both of them made comments during the conversation that pointed to an embrace of a high touch approach.  So I thank them for that and hope to get a few more folks on the bandwagon soon.  If anyone out there would like to chat about this further - face to face or virtually, I'd be delighted to do so. 

No comments: