If something is important to me, I would work like a dog to achieve it if I had to.  However, if I could achieve this important outcome without hard work, why would I work hard for it, even though I would accomplish much more by working hard for it?  That’s just simple human nature, and it encapsulates the major harm to our educational system done by grade inflation.

Students want high grades for many different reasons (e.g., admission to post-secondary professional schools, jobs, a sense of accomplishment).  When high grades are difficult to achieve, students will work exceptionally hard in their effort to achieve them.  However, why would students bust their butts studying in a class where everybody gets A’s regardless of their effort?  The answer is they won’t, and it turns out, that’s what’s happening.  

The most pernicious effect of grade inflation on education is to cause a substantial diminution of student effort in their coursework.  Students now get better grades for much less effort.  The corollary of this is that students are learning much less now from the same course material than when average grades were much lower.  And there’s data to prove this!!

Philip Babcock was an economist who wrote some of the best papers analyzing data on the effects of grade inflation on student effort in the past decade.  I would strongly encourage anyone interested in this subject to seek out his full output.  Tragically, he died in 2012 while he was an Assistant Professor at the University of California, Santa Barbara.  However, before his death, he left two important papers that considered the relationship between student effort and grades.

In one published in Economic Inquiry in 2010 (the preprint of the paper can be obtained here from the UCSB archive), he analyzed the relationship of self-reported hours worked outside of class on course work and expected grades (i.e., what the student expected to receive, but reported before the students knew what their grades were).  I quote from the abstract of that paper:

Results indicate that average study time would be about 50% lower in a class in which the average expected grade was an “A” than in the same course taught by the same instructor in which students expected a “C.”

Think about that number!  His regression analyses indicate that students in classes where they expect to get an A study half as much as students in classes where they expected to receive a C.  Thus, simply based on the grades that the average student receives in the class, the work effort on the material for all students in the class could double depending on how easy or hard a grader you are.  

Let’s say you went into a class one year, and on the first day you said, “I am going to set the average grade for this class to a C” and then you simply teach the class as you always do but grade assignments to a C average.  Then in the next semester, you teach the identical class, but on the first day you say “I am going to set the average grade for this class to an A”, and then teach exactly the same class but grade all assignments to an A average.  Which class do you think students would get a better education in, and by “get a better education” I mean know more, be better able to apply what they learned in the class to new situations, and have more skill with the material in the class?  Based on human nature alone, the answer is obvious.  

I see the same relationship in analyses I have done of the self-reported data from Dartmouth students when they complete their course evaluations at the end of each term.  Dartmouth has been collecting these data since 2006 from students, and we have almost 250,000 responses so far.  In my presentation to the Dartmouth faculty last Monday, I showed Babcock’s results and then the Dartmouth student data, which are very consistent with one another.  Dartmouth has not yet given our committee permission to post the full set of analyses we completed – hopefully, they will very soon.  However, many of the figures I presented can be found on dartblog.com, and I link to the pertinent one here.  Dartmouth does not ask the questions in a way that permits me to analyze our data as thoroughly as Babcock did for the UCSB data, but the very same relationship is apparent.  The percentage of students that self-report working 1-10 hours per week outside of class on coursework declines as their expected grade declines, and the percentage of students who self-report spending more than 15 hours per week outside of class on the coursework increases as their expected grade decreases. 

In the Dartmouth data, one can also see this effort taken to its extreme.  The “NRO” grade was a Dartmouth innovation introduced in the 1966-67 academic year, apparently in an attempt to curb grade inflation even then.  NRO stands for Non-Recording Option.  Students taking a course NRO specify before the term begins what minimum grade they would like to achieve in the class.  If the student’s final grade is at least their minimum, that grade is reported on their transcript; but if the final grade is below this minimum, then the student is graded Pass/Fail.  In the histogram above, the green bars are for students who passed the class but did not meet their minimum grade (NRO on their transcript).  From the histogram, ~14% of these students self-report working <1 hour per week on this class, and ~40% self-report working 1-5 hours per week.  These two percentages are by far the highest in those two categories.  

Another paper published by Babcock and Mindy Marks (University of California, Riverside) in 2010 in The Review of Economics and Statistics (preprint archived at The National Bureau of Economic Research is here) takes a longer and national view of this relationship.  They report their analysis of a compendium of data from four national surveys of United States college students.   Again, instead of summarizing, I quote from their abstract:

Full-time students allocated 40 hours per week toward class and studying in 1961, whereas by 2003, they were investing about 27 hours per week. Declines were extremely broad based and are not easily accounted for by framing effects, work or major choices, or compositional changes in students or schools. We conclude that there have been substantial changes over time in the quantity or manner of human capital production on college campuses.

In fact, in the paper they report specifically a 40% drop in time spent working outside of class on coursework from 1961 to 2003.  The Dartmouth data shows a linear increase in the grades given to students over this exact period, and to today, and based on information in the press, Dartmouth is completely unexceptional in the rate and scope of grade inflation. 

Here is exactly why the US standing in international academic achievement in many different categories has been falling over this same period.  University student effort on coursework has declined for at least the past 50 years as the grades given to them over that same 50-year period have increased linearly.

If you also talk to many students today about what they value about college, you find out that it’s not their coursework.  That is certainly true among a segment of the student body at Dartmouth.  For example, the student Editorial Board for The Dartmouth published an Opinion piece last Friday in which they made the argument that extracurricular activities were more important to them that their courses.

The goal of any liberal arts education is to challenge students to think critically, take risks and grow intellectually alongside their peers. Assessments and homework are just one component of such an education. Increasing the intensity risks undermining this goal by limiting time available for personal care, extracurricular pursuits and social bonding — all of which are undeniably important to student well-being and overall achievement.

Their argument is that if you increase the academic rigor of courses, they won’t have enough personal time or time to bond.  My question to those students is what does any of that have to do with why you are at Dartmouth or attending any university?

They also equate experiential learning/study abroad/innovative teaching with a lack of academic rigor:

Traveling abroad is, of course, an exceptional example — a focus on engaging and consistent learning is possible in Hanover as well. Even less resource-intensive changes, such as including collaborative or creative projects, oral assessments and small discussion sections in one’s curriculum, go a long way toward teaching students both specific course content and broader intellectual skills. These skills, not the mastery of content necessary to receive an A in any given course, make a Dartmouth education valuable.

These are all ways that Dartmouth faculty continue to innovate in our curriculum, but course innovation goes hand in hand with increasing academic rigor, not against it.  

Luckily, many of Dartmouth students with which our committee discussed our proposal for increasing academic rigor decidedly do not have the same opinions of these issues as the student Editorial Board of The Dartmouth.  

The thing that our committee continually stresses to everyone is that the only way to fix grade inflation is to 

give all high-performing students high grades, all intermediate-performing students intermediate grades, and all low-performing students low grades.

Our guiding principle in everything we propose is that every student who develops excellent mastery of the material and educational goals of a course should get an A for that course.  

The reason previous attempts to curb grade inflation have failed (e.g., see here and here) is because they destroyed the incentive structure for students to achieve.  Caps and quotas on the fraction of A’s destroy this incentive completely: the carrot in this incentive structure.  However, grade inflation removes the incentive to actually work for that A: the stick in the incentive structure.  You can’t reinstitute the stick without also having the carrot in place as well.  Who wants to just get beaten with a stick?

So our plan empowers “a notion radical in its simplicity” (to quote Joe Asch, Dartmouth 1979), keep the “carrot” for students of everyone who deserves an A should get an A, but also reconstitute the “stick” of getting low grades when performance warrants such a grade.  It’s actually pretty simple.  We hope that reconstituting this incentive structure for students will lead to massive increases in the overall educational benefits to them.

The only conclusion that one can draw from these many lines of data is that fixing grade inflation by returning to a system in which low performing students receive low grades will be the motivator that will return student engagement and effort with their coursework to acceptable levels.  This is the fundamental goal we have for all our work.