Blog Post

How to Get Out of Grading

Inside Higher Ed has also picked up the "How to Crowdsource Grading" blog and features a very nice and thoughtful follow-up by the always thoughtful Scott Jaschik:


It's called "How To Get out of Grading" but isn't remotely as cynical as the title suggests.  As anyone who does this kind of peer review knows, it takes more time, not less.   And I can't wait.   I'm getting flooded with comments; I'm told it is the most read piece in Chronicle of Higher Education this month so far; some people are positive, some negative (vehemently in some cases), and some just curious.


Everyone:  it's an experiment!  I promise to report back.  And I am delighted by the interest.  People can be very cynical about college teachers and teachers in general but in my experience most of us continually try and try harder to keep learning meaningful and exciting, sometimes against great odds.  All of this interest signals to me that we all are concerned with pedagogy and want assessment--but realize that our forms of assessment, now, are impoverished.  We can do better. 


I would be especially interested in hearing from those who have tried new forms of assessment.  What are they?  How did they work?


And here is the link to the Chronicle of Higher Education's article by Erica Hendry:


And, for anyone who is really interested in this coverage, here's the Raleigh News and Observer piece:



Hello Cathy,

I enjoyed reading your previous article on "How to Crowdsource Grading" very much. I am a software professional who often teaches programming at a local college. I enjoy teaching and interacting with the students, but grading is a different beast altogether :-)

I agree with the title of this article. I have always felt that having the grade a class reduces the quality of my teaching. Since there are supporters as well as detractors fir this school of thought (perhaps both with good reason), I will give my reasons for disliking grading:

There are two aspects of programming

1. The syntax of a programming langauge, which are fixed and can be called facts.

2. The art of programming, which is a bit subjective.

It is easy to test students on their knowledge of the facts, but then we do not really need to have a course just for the facts. They can be learnt by reading any good book on Java programming. A course and a tecaher is needed to teach the "art of programming", and this is difficult to evaluate in a fair way.

Because it is difficult to grade the "good programming practices" part, I cannot make it part of the tests, and because I cannot make it part of the tests, I have to debate with myself, how much time I should dedicate to it in the classroom.

The right way to run a programming course (at least in my opinion) is to run the course as an open lab. The students are expected to read and understand the language syntax before they come to class, and in the class they are expected to work on actual programming assignments. The teacher should move among the students answering their questions and helping them with the programming. This way students would learn a lot more about prpogramming. Students should also be encouraged to collaborate with eac other as they work.

I feel that such a course will help students learn programming far better than traditional courses. However, in such a setting, it becomes very difficult to grade students in a fair way. I cannot implement this type of teaching in my classes mainly because it is difficult to grade. I am forced to teach in a less effective way simply because it is easier to grade students.

I think we should ask a very fundamental question. Why do we grade students? Is it to give feedback to the students and help them understand where they need to fill in the gaps of their knowledge, or do we grade students so the corporate world can use grades as a filter? I almost feel that grading fosters the kind of thinking which we should actually teach students to stay away from: "studying only for grades and working only for money".

If we establish the purpose of grades is to help students, then we can give them feedback without the grades and let the students' publicly available portfolio speak for their skills.

Thank you for trying to implement such an innovative method for grading, in your course. I hope it is successful.



Parag Shah


Dear Parag Shah,  Thanks so much for this illuminating comment.  Yes, it's partly from my experiences with various certification programs--in labs, engineering, medicine--that I began to think about how un-nimble, in intellectual terms, assigning a grade is for really evaluating learning.  Some commenters have suggested that my system fails to prepare students for "the real world."  But, in fact, there are no midterms and finals in the real world.  In the workplace, one is constantly "tested" by one's ability to perform as an individual and as part of a team, by one's ability to solve problems and by the respect one has from one's peers as well as from one's supervisor for one's ability to 'deliver', whether that be an innovative solution, a timely response, an intervention, or a workplace correction that saves the day.  


Grading itself is not standardized and not timeless.  Historians dispute who gave the first "grade" but most put it at the late eighteenth or early nineteenth century, with the beginnings of mass literacy, the very beginnings of a middle-class culture, and the first whisperings that a "gentleman's" liberal education was not the only way.  With the nineteenth century, we see the beginnings of professional schools and an expanded pool of people (including women) entering colleges and professional schools and certification becomes important.   Grades and standardized testing really gets going in the late nineteenth century, with Galton, and also with the Machine Age and good old Taylor taking out his stop watch to measure the efficiency of this man carrying a wheel barrow with so much weight or that one performing an individuated task on an assembly line.  Standardized metrics for segmented performance and standardized tests for discrete measurement of specific abilities go hand in hand.


So what is the metric for assessing a digital age when production and consumption are fused, when people who never even met one another can collaborate together to create the world's largest encyclopedia or algorithms that solve major digital problems (identification, preference, selection, distribution), and so forth?   How do we assess group performance that is anonymous, collaborative, customizing, and often undertaken simply for the sake of learning more or doing more or creating more and more interesting strategy (as in MMOPG's)?   If the Machine Age made grading, IQ tests, standardized testing, and other metrics standard for a form of discretely individualized assessment, what is the ideal way of evaluating "smart mobs," or "the wisdom of crowds," or "here comes everybody" participatory learning?


I don't have the answer but I know it is a great question.  I'm engaging in this experiment in order to see what, together, in the manner of this Information Age, I and my students can think about as a new, relevant, important way of assessing what the digital world does and means and what any individual contribution might mean.   "This Is Your Brain on the Internet" is about precisely rethinking those Machine Age categories---and reassessing them!


My first assignment for this course will be reading all the comments on this form of assessment and assessing them.  Yours is a very productive one that my students will learn from.  Thank you!