Who's tried Ungrading in STEM courses?

[vela]

I'm curious if anyone has tried "ungrading" in their STEM courses and if so, how it worked out. The idea is to get away from the using points to determine a student's grade and use different types of assessment that better motivate students to learn.

https://www.jessestommel.com/why-i-dont-grade/

https://www.chemedx.org/blog/ungrading-what-it-and-why-should-we-use-it

My interest arises from my experiences since classes went remote because of the pandemic. Like many other instructors, I saw the mysterious increase in performance by many students on exams (as well as obvious signs of cheating in some cases). To reduce the incentive to cheat, I replaced most of these high-stakes assignments with low-stakes weekly problems, where students had to write up a solution where they had to identify the relevant physical concepts, explain their problem-solving strategy, and finally solve the problem. It wasn't enough to just write down a bunch of math, which they could easily find on Chegg or somewhere else on the internet; they actually had to articulate the reasoning involved. I have to say I was pleasantly surprised by the quality of the write-ups from some of my students.

There were some problems, however. The main thing was assessment. I developed a rubric, but then it would sometimes end up resulting in a grade I didn't feel accurately reflected the quality of the work. Over time, I've modified the rubric, but I've never been happy with the results. This semester, I'm considering just giving them scores of "satisfactory," "needs revision," and "not submitted," and record audio feedback on what I thought they did well, what could use improvement, etc. I'm still thinking about how to translate these results into a letter grade that I have to assign at the end of the semester.

Anyway, I would love to hear any comments or idea, tips, and about anyone's (student or faculty) experiences with these types of assessments.

 

[kuruman]

The institution where I taught required me to submit grades on the standard scale A = Excellent, B = Very Good, etc. with pluses and minuses. I had no choice in the matter. I was also required to outline my grading procedures in my published syllabus which was considered a binding document. Needless to say I was in complete control of how to translate a student's performance over an entire semester to one letter grade. I approached grading differently in intermediate physics courses (junior-senior level mostly for majors) from (service) introductory physics courses.

At the intermediate level, I taught small classes (N ≈ 16) that allowed me to figure out what each student could do in a matter of 3 weeks. The homework in theses courses gave me a reading of the students' skills and abilities. In class I asked questions to the entire class and saw who was eager to answer them. Three in-class tests (the final was one of them) gave me an idea of how well the students were able to pull everything together. My final grade was weighed 1/3 homework, 1/3 class participation and 1/3 in class-tests. The class participation score was the least objective but it allowed me to cast a "vote of confidence" to students who were eager to learn but had to overcome obstacles. I have an instructive story about this.

A couple of hours before an in-class exam a student walks into my office wanting to know how to do a particular problem from the homework. As luck would have it, this was one of the exam problems. There was no way the student could have gotten hold of the exam because I had just finished putting together the original and was about to print the copies. What should I do?

I guided the student by asking, PF style, one leading question after another, never giving anything away, until she got to the end successfully. When she left I was sure she would ace this problem. Well, she bombed it. When I returned her paper I asked her, "How could you miss that?" She said "Because of the pressure of knowing that I could do it, I froze and I couldn't do it." What should I do now? This student, under relaxed conditions, showed that she could do the problem. Under the double stress of the in-class test and the "I better know how to do this" pressure, she folded. So I calculated her final grade for the course as if she had gotten this problem correctly on the test. My rationale was that the information that the student could do that problem was valid even though it was not obtained via the in-class test. Giving her a lower grade because she bombed the question on the test would be punitive, not an assessment of how well she can do physics.

Introductory level courses were a different beast altogether to me. Their relatively large number of students (120 < N < 180) obviated meaningful class participation. Homework was delivered and graded by web-based algorithms and only in-lecture hall tests with word problems gave some insight into how students think about solving a problem.

This post has already grown long enough. I will not go through my introductory course grading ideas and experiences here. I will come back for that later with more stories to tell.

 

[gleem]

My rationale was that the information that the student could do that problem was valid even though it was not obtained via the in-class test. Giving her a lower grade because she bombed the question on the test would be punitive, not an assessment of how well she can do physics.

But couldn't have other students froze under the pressure too? Was it fair to them? I don't want to seem hard hearted, but I don't think I could give her full credit.

 

[berkeman]

But couldn't have other students froze under the pressure too? Was it fair to them? I don't want to seem hard hearted, but I don't think I could give her full credit.

Yeah, that was my initial thought as well, but I trust kuruman's judgement so I let it go. @kuruman -- was this the hardest bonus question on the test, or a straightforward one in the middle of the test?

 

[vela]

But couldn't have other students froze under the pressure too? Was it fair to them? I don't want to seem hard hearted, but I don't think I could give her full credit.

It's hard to argue that it was fair to the other students, but at the same time, is it fair to penalize her for not being a good test taker, even though she understood the material? It's issues like this that has encouraged some teachers to look into different methods to assess students.

 

[berkeman]

It's almost like an in-person (or Teams) individual exam would be more accurate. For small class sizes that could work. For larger classes, it would take a lot of TAs to make it work...

In my in-person undergrad physics classes back in the mid-70s, I was most impressed by the professors who posted the bell curves for each exam on their billboards. We all gathered to see our scores and how they compared to the rest of the bell curve...

 

[kuruman]

But couldn't have other students froze under the pressure too? Was it fair to them? I don't want to seem hard hearted, but I don't think I could give her full credit.

Consider for a moment why you think that fairness (or lack thereof) is an issue here. A good grade is not a reward for a job well done and a bad grade is not punishment for falling short of expectations. The grades I gave were my honest, professional and informed assessment of how well a student has been able to do physics. The homework, tests, etc. that I asked everybody to do to were the minimum necessary to inform my judgment. If it so happens that I come across additional information, I fold it in my assessment as another data point. As long as I do this for all students and do not exclude anyone, I am not being unfair. I often told my students, "The exams are not designed to test what you cannot do but what you can do. Give me an excuse to give you a good grade and I will."

Yeah, that was my initial thought as well, but I trust kuruman's judgement so I let it go. @kuruman -- was this the hardest bonus question on the test, or a straightforward one in the middle of the test?

It was one of three EM problems of comparable difficulty that I asked students to do in 90 minutes. Each problem had three to four parts of increasing difficulty.

Thank you, @berkeman, for trusting my judgment. I grappled a lot with the ethics and fairness issues in this particular case which I resolved by considering the award of a grade as providing an interpretation of a series of physical measurements. The events of the case that I described here occurred are than 30 years ago but they are still vivid in my mind because they shaped my attitude towards grades from then on.

 

[gleem]

That was my opinion I was not trying to be judgmental. Yes, I understand your reasoning, but still other students could claim the same reason for underperforming. (We hear a lot about test anxiety in this forum.) Should it have been incumbent on you to have asked every student who did poorly on that question why they did?

Yes, I know any student could have come to you as ask how to do that problem and maybe some did. But still this student had a decided edge.

The grades I gave were my honest, professional and informed assessment of how well a student has been able to do physics.

Subjective. Doesn't this burden you on making decision on the student future performance. What do you do about other questions that this student does not do well on "knowing" she does not take tests well?In thinking about this situation, I would have replaced that question, having a few hours before the test.

Failing to have been unable to do this I think I would have left the grade the same, kept an eye on the student subsequent performance.

You are not punishing a student if they do not meet standards of performance. Your tests should be an objective measure of the knowledge/performance that you expect. We are all required to "stand and deliver" throughout our lives. Sometimes we don't. Most of the time we just pick ourselves up and move on. Some people don't or can't.

 

[kuruman]

Yes, I know any student could have come to you as ask how to do that problem and maybe some did. But still this student had a decided edge.

I agree with you here and I see nothing wrong. By asking me that question before the test, she got the same edge that students here at PF acquire when they come to us with homework questions. It's called learning. Students who learn have a decided advantage over students who don't.

Subjective. Doesn't this burden you on making decision on the student future performance. What do you do about other questions that this student does not do well on "knowing" she does not take tests well?

Of course it's subjective. I, not a machine, evaluated this particular student's overall performance to give a final grade on the basis of all the pieces of information that I gathered about her over the entire semester. Her missing that particular problem was inconsistent with what else I knew about her and her and with her performance on the other two exam problems which was fine. In a small classroom it is easy to figure out who is a good student and who is a slacker.

Failing to have been unable to do this I think I would have left the grade the same, kept an eye on the student subsequent performance.

The student's subsequent performance was fine. After graduation she got her M.D. and specialized in psychiatry. Last I heard of her she went to practice in the Southwest after learning how to fly plane so that she can move quickly from one reservation to another and tend to her Native American patients.

 

[berkeman]

The student's subsequent performance was fine. After graduation she got her M.D. and specialized in psychiatry. Last I heard of her she went to practice in the Southwest after learning how to fly plane so that she can move quickly from one reservation to another and tend to her Native American patients.

Very nice!

 

[mathwonk]

my practice was always to try to make grades mean exactly what i had said they would mean when I defined them on the syllabus: so much for homework so much for exam, so much for presentation. that way, i felt that at least anyone afterward could see that grade and know exactly what it meant, and every student was measured with the same stick. inevitably this did not necessarily reflect exactly how much someone "knew" under ideal condtions.

because of the challenges of creating better measures over the decades, i tried many other methods, including once just guaranteeing everyone a good grade who hung in there faithfully to the end and did what was asked, maybe even once guaranteeing an A, don't remember now for sure. In those no stress situations, good students did wonderfully well, feeling an obligation to give their all and earn the grade, but with slacker students it only allowed them to deceive themselves as to how they had done. In my own education, one of my professors guaranteed at least a B to anyone who did "not fall flat on his or her face", on the exam. That same professor however had scared off at least one student the first day by announcing that the course was "hard, hard, hard, making strong men weep and women cry". At that point one young man got up and walked out. It was only months later after we had endured the course that he made the guaranteed grade announcement. This was math 55 at Harvard, arguably the "hardest course in america". https://en.wikipedia.org/wiki/Math_55

but i have no advice or criticism for how anyone else grades his or her course. the grade should be whatever the instructor thinks it should be in my view, in consultation with his/her conscience (and syllabus).

 

[symbolipoint]

Give me an excuse to give you a good grade and I will."

Is this the same as, "Give me a justifiable reason to give you a good grade, and I will." ?

 

[kuruman]

I was most impressed by the professors who posted the bell curves for each exam on their billboards.

Part II (Grading Introductory Courses)

Since I was required to submit a letter grade, I felt obliged to make that letter grade as meaningful as possible in the introductory courses where the size of the class prevented me from figuring out what each individual student could and could not do. I never succumbed to the allure of bell curves and never used them. If I understand correctly, the assumption is that, in a large class, the parent distribution of grades is a bell shaped curve. One then fits the actual distribution obtained after a test to the parent distribution so that the relative percentages of A's, B's etc. are more or less the same from one test to the other. I think the intention behind this procedure is to give an out to the instructor who unintentionally designs a particularly difficult exam only to discover that the highest score is a 65% obtained by only one student. Saved by the curve!

The problem I have with "curving tests" is that it leads one down the slippery slope of grade inflation. As a measurement, a curved grade is not an independent assessment of a particular student's performance because it depends on the student average, i.e. how everybody else in that particular class did. Thus, if a class one year consists primarily of smart people and next year it's all dummies, and both years the grade distributions are the same, how is that a fair and accurate assessment of the abilities of either group? This is what I did with my tests in introductory courses.

Before printing the exam, I actually took it and timed myself. Sometimes I found that a particular task took longer than I thought which meant that it had to be modified or deleted. If I completed the exam in about 16 minutes, I deemed that it was not too long for a 50-minute in class test under the assumption that the average student would need a factor of 3 more time.

After collecting the papers, I sat down with the TAs and set grading criteria trying to cover all eventualities and common mistakes. Then we graded and tallied the scores for each paper. Now the interesting part: what do these scores mean in terms of the standard scale A = Excellent, B = Very Good, C = Fair, D = Poor, F = Failure? That's where my judgment as the course instructor comes in. The numerical standard scores (SS) was F < 60, 60 ≤ D < 70, 70 ≤ C < 80, 80 ≤ B < 90, 90 ≤ A ≤ 100. Pluses and minuses were spaced 10/3 points apart within a letter interval.

Now comes the task of converting the raw scores (RS) into standard SS scores. To do that, I looked at the exam and the grading criteria and asked myself the question, "What is the minimum raw score that a student must get in order to barely pass this exam? How forgiving do I want to be?" In most cases the answer was about 40% of the maximum raw score. I called that value RS₁. I then asked myself, "What is minimum raw score that a student must get to cross the line from B+ to A-? How generous do I want to be?" I called that value RS₂. Having the two values, I constructed a straight line passing through the points {RS₁, 60} and {RS₂, 90} and used that to convert raw scores into standard scores for each student. Here is why I think this method is superior to "curving"

• The grade of anyone individual student is decoupled from the others. If the entire class deserves A's, the entire class will get A's; if the entire class deserves D's, the entire class will get D's.
• By setting the RS points at the same values year after year not only I guard against inflation, but also I can compare class quality from one year to the other.
• I have the flexibility to give exams with different maximum raw scores, say 75 on one and 85 on another. This allows me to throw out a question that makes the exam too long without having to readjust the maximum score to the usual value of 100.
• The straight line SS = a (RS) + b has a positive intercept around 30 SS points. This means that if a student collects only a handful of raw score points, he/she has 30-40 standard points in the bank. Still below 60 and a failure, but not as catastrophic as the Gaussian tail would engender. This gives hope to the student for a comeback.

I will stop here. I intend to continue with a third installment in which I will start with an astonishing experience and then finish with turbo averaging, what it is and the event that led me to implement it.

 

[vela]

As long as I do this for all students and do not exclude anyone, I am not being unfair.

One complication I see here is that it might not occur to some students that these opportunities exist, or they are just not the type to take advantage of them, perhaps for personal or cultural reasons. For example, when I had a class working on a problem, it was like pulling teeth to get many of my Asian students to talk to people outside of their peer group, and most would never ask me for help if they were stuck. If they had been in your class, these students would never have gone to your office with a question.

Your tests should be an objective measure of the knowledge/performance that you expect.

The assumption here is that the tests are a good measure. The point of my thread is asking if there are different and better ways to assess students, ones which perhaps minimize or eliminate factors like test anxiety and give an incentive to students to engage with the material more meaningfully, rather than asking "is this going to be on the test?"

 

[vela]

Because of the challenges of creating better measures over the decades, i tried many other methods, including once just guaranteeing everyone a good grade who hung in there faithfully to the end and did what was asked, maybe even once guaranteeing an A, don't remember now for sure. In those no stress situations, good students did wonderfully well, feeling an obligation to give their all and earn the grade, but with slacker students it only allowed them to deceive themselves as to how they had done.

I'm curious how you determined whether a student had done what was asked. What was the difference between what a good student did and what a slacker did?

 

[gmax137]

Part II (Grading Introductory Courses)

Fascinating. I have never been involved in scoring tests nor assigning grades. Reading this makes me wonder how professors in other disciplines assign grades. Subjects like, say, english composition or french lit? Those subjects are concerned with critical reading/thinking, and clear communication of well-thought-out ideas. I can't imaging how the professor there decides, "this essay is 74 points, that one is 75..." More likely the professors "know an 'A' essay when I read one" and the same for other grades?

 

[symbolipoint]

Fascinating. I have never been involved in scoring tests nor assigning grades. Reading this makes me wonder how professors in other disciplines assign grades. Subjects like, say, english composition or french lit? Those subjects are concerned with critical reading/thinking, and clear communication of well-thought-out ideas. I can't imaging how the professor there decides, "this essay is 74 points, that one is 75..." More likely the professors "know an 'A' essay when I read one" and the same for other grades?

No matter the complication, determining a score is easier than you think it is. Ponder on the idea of *"rubric". And also the person doing the grading or assessing must use his own judgement. ( I guess I should try to create an example...?.)* Note that @vela also used the term, "rubric" in the initial post.

 

[sysprog]

The word 'rubric' referred in Middle English to a section written in red.

 

[vela]

Fascinating. I have never been involved in scoring tests nor assigning grades. Reading this makes me wonder how professors in other disciplines assign grades. Subjects like, say, english composition or french lit? Those subjects are concerned with critical reading/thinking, and clear communication of well-thought-out ideas. I can't imagine how the professor there decides, "this essay is 74 points, that one is 75..." More likely the professors "know an 'A' essay when I read one" and the same for other grades?

One question that comes to mind is what's the difference between a 74 and 75? Is it significant? What does such a score communicate to the student?

* Note that @vela also used the term, "rubric" in the initial post.

You also might note I am going to stop using one and try a different form of feedback than giving points.

 

[kuruman]

The assumption here is that the tests are a good measure. The point of my thread is asking if there are different and better ways to assess students, ones which perhaps minimize or eliminate factors like test anxiety and give an incentive to students to engage with the material more meaningfully, rather than asking "is this going to be on the test?"

How much time per student can you allocate for assessment? Nothing beats the oral exam if you want to know how much a student knows and can do. It is manageable in a small class of, say, 20 students. Assuming 20 minutes per students, that's about 7 hours. It would take you that much to compose, print, administer and grade a written exam. However, I think it is unnecessary for a small class. In my small classes I told students that "class participation" would constitute 1/3 of their grade. My definition of class participation to them was, "If you ask questions during class, that's class participation. From time to time I will ask questions directed to the entire class. You don't have to raise your hand to answer them, but if you do, that's class participation whether you answered correctly or not."

As you can see class participation served as a substitute oral exam, spread over an entire semester. If you are wondering "what about timid students or international students who have cultural barriers that prevent them from asking questions", I would say there is only so much one can do. One cannot grab a student by the ear and say, "Hey, come to my office and ask your questions."

The time per student for individualized assessment in a large class is ridiculously low. When I taught 120 students in a large class, I had set aside 4 hours per week for office hours. If everybody came, I would have only 2 minutes to spend with each student. Doubling the office hours, would bring the time per student to a whopping 4 minutes. Something has to be sacrificed and that is individualized assessment. You haven't mentioned the size of your class, @vela; it makes a difference.

 

[gleem]

One question that comes to mind is what's the difference between a 74 and 75? Is it significant? What does such a score communicate to the student?

Another question. What is the difference between 9.58 seconds and 9.59 seconds? A gold medal and a world record. I know that doesn't answer your question, but so many things come down to something seemingly insignificant but have dramatic consequences. A hop in a dismount in a gymnastics competition, making or not making the team.

Haven't we all dealt with grade definitions?

 

[kuruman]

One question that comes to mind is what's the difference between a 74 and 75? Is it significant? What does such a score communicate to the student?

In my book there is no difference, zero, zilch. That's because I was required to submit letter grades with pluses and minus which necessitates a conversion from decimal to alphanumeric. If the bins in the alphanumeric histogram are wider than the bins in the numeric histogram, then two numerical grades may very well fall within the same alphanumeric bin and the two students get the same letter grade. The nuance of the difference in performance is lost.

For grade G, the C range was $70 \leq G < 80.$ I subdivided the range in bins of equal widths:
$70 \leq \text{C -} < 70+\frac{10}{3}$
$70+\frac{10}{3} \leq \text{C} < 70+\frac{20}{3}$
$70+\frac{20}{3} \leq \text{C +} < 80$.

A 74 would be a C and so would a 75. The real problem is at the boundaries. For example, what grade do I assign to a score of 79.999? An Excel algorithm on Excel would return C+ and I could leave it at that if I didn't't know any better. But I do know better. I know that grading a paper, even when the grading criteria are exquisitely detailed, is a measurement with built-in uncertainty. If "drawing a correct free body diagram with all the forces labeled" is worth 3 points and the paper shows only 2 instead of the correct 3 with one in the wrong direction how many points do I assign? One may be too low and 2 too high. At this point, although the criterion is objective, I have to make a subjective decision. Whatever the decision, there is an uncertainty of ±1 in this particular problem in this particular test that propagates all the way to the final numerical answer together with its fellow uncertainties from this and all the other pieces of work that the student turned in. This means that I have to estimate an overall uncertainty, another subjective task. To cut a long story short, I decided that 0.3 points in favor of the student would be appropriate and generous to account for all uncertainties. That is, 79.71 is a B- and 79.7 is a C+. In practice, 3%-4% of the students were given this benefit of the doubt across all boundaries.

 

[vela]

Another question. What is the difference between 9.58 seconds and 9.59 seconds? A gold medal and a world record. I know that doesn't answer your question, but so many things come down to something seemingly insignificant but have dramatic consequences. A hop in a dismount in a gymnastics competition, making or not making the team.

Haven't we all dealt with grade definitions?

In a race, the goal is to reach the finish line in the least amount of time. It makes sense to measure and compare times. In school, the ultimate goal to not to get students to score 100 on an exam or paper; it's to get students to learn. Alfie Kohn has argued giving grades works against that goal.

In the 1980s and ‘90s, educational psychologists systematically studied the effects of grades. As I’ve reported elsewhere (Kohn, 1999a, 1999b, 1999c), when students from elementary school to college who are led to focus on grades are compared with those who aren’t, the results support three robust conclusions:

* Grades tend to diminish students’ interest in whatever they’re learning. A “grading orientation” and a “learning orientation” have been shown to be inversely related and, as far as I can tell, every study that has ever investigated the impact on intrinsic motivation of receiving grades (or instructions that emphasize the importance of getting good grades) has found a negative effect.

* Grades create a preference for the easiest possible task. Impress upon students that what they’re doing will count toward their grade, and their response will likely be to avoid taking any unnecessary intellectual risks. They’ll choose a shorter book, or a project on a familiar topic, in order to minimize the chance of doing poorly — not because they’re “unmotivated” but because they’re rational. They’re responding to adults who, by telling them the goal is to get a good mark, have sent the message that success matters more than learning.

* Grades tend to reduce the quality of students’ thinking. They may skim books for what they’ll “need to know.” They’re less likely to wonder, say, “How can we be sure that’s true?” than to ask “Is this going to be on the test?” In one experiment, students told they’d be graded on how well they learned a social studies lesson had more trouble understanding the main point of the text than did students who were told that no grades would be involved. Even on a measure of rote recall, the graded group remembered fewer facts a week later (Grolnick and Ryan, 1987).

https://www.alfiekohn.org/article/case-grades/

 

[Andy Resnick]

I'm curious if anyone has tried "ungrading" in their STEM courses and if so, how it worked out. The idea is to get away from the using points to determine a student's grade and use different types of assessment that better motivate students to learn.

[...]

Anyway, I would love to hear any comments or idea, tips, and about anyone's (student or faculty) experiences with these types of assessments.

There's a lot in this video that I am skeptical of- and to be fair, it's not clear if Dr. Stommel is primarily discussing K-12 or higher education. However, I don't think his approach works (at least for STEM), and I base that on our experience with teaching/grading during the pandemic.

For the past several semesters, ending with this one, my institution has permitted students to choose whether or not a class they took should be graded as pass-fail: this is a per-student, per-course decision implemented by the Registrar (the instructor, if they chose to grade, assigned a grade as normal and upon petition by a student, the registrar replaced the grade with a pass/fail grade). Many instructors simply decided to grade pass-fail en bloc, but my point is that my institution experimented with ungrading during the pandemic.

It has been a disaster. Students were not held accountable for the material and so multi-semester sequences (Physics I and II, for example) resulted in those students being unusually poorly prepared for the following course. Higher ed faculty are going to be dealing with pandemic-related K-12 learning deficiencies for years, and I'm not sure how that is going to play out. Ungrading will just exacerbate this problem because it removes a motivating factor for students to push themselves.

Interestingly, students have been the ones to push administration to bring grading back. My suspicion is that it's the high-achieving students who pushed the hardest.

Dr. Stommel doesn't spend much time addressing the fact that many students must take classes they don't want to (e.g. required classes). His claim that grades force students to do things they don't want to do is an odd claim, what does he propose to do about required courses?

Regarding different types of assessment, I'm all in favor of that: low-stakes formative assessments, group projects, essays, etc. etc. At semester's end, though, I have to assign a letter grade to each student.

Regarding the comments about grading on a curve, I don't do that- but not for the reasons given on this thread. I don't grade on a curve because then, by definition, some of my students MUST fail regardless how hard they work and try.

[note added]: another example of 'ungrading' is when higher ed institutions drop requirements for the SAT/ACT/GRE/TOEFL/etc; we have done some of this (dropping GRE requirement) due to the pandemic, I'm not sure what the overall impact of this is.

 

[kuruman]

Regarding the comments about grading on a curve, I don't do that- but not for the reasons given on this thread. I don't grade on a curve because then, by definition, some of my students MUST fail regardless how hard they work and try.

That's exactly the reason why I think that grading on a curve is evil. In my scheme if the entire class deserves to fail or if the entire class deserves an A, so be it. Grading on a curve makes this impossible. I should add that neither situation materialized in any of my classes not because it was impossible but because it was improbable.

The astonishing experience with grading that I mentioned earlier was this. After grading and scaling the second exam in a large course I took the class average to compare with the first. Much to my astonishment, the second average was two one-hundredths of a point of the first exam average. I had never seen anything like this before so I did some more digging to see how individual students fared. Students whose second score was within ±5% of the first score were labeled as "no change in performance." Students outside that range were labeled "worse performance" (low end) or "better performance" (high end) than last time. Well, the tally showed 20% no "change", 40% "worse" and 40% "better". My interpretation was that some students who did well the first time around slacked off and rested on their laurels as it were, while others who did not do so well got scared and took measures to do better. As luck would have it, the numbers in each group were about equal and the 40% above and below swapped places relative to average.

If my interpretation is correct, then one can argue that it shows that grades serve as an incentive to do better and, simultaneously, as a ustification to slack off. My opinion is that it shows that some students have a strong innate drive to always do better than before while others do not. If the drive isn't there, it cannot be injected externally by changing methods of assessment. One can inspire students to learn only if the students have the correct attitude towards learning which not all students have. Those of us who have had non-traditional students (aged older than 25) in their classes know what I mean. One of their main attributes is that they don't give a rat's behind about grades. They will do whatever is necessary to learn because they are in school to do just that and improve themselves. I am unconvinced that all 18 year olds who go to college are there to do just that.

Example: Student majoring in physics quits after a year of classes with a D- average. He enlists in the Marines and spends 6 years with them. He comes back to us, married with a child, still a physics major. He aces all his classes (by his own admission he was "partying too much" the first time around.) He graduates with honors and continues on for a Ph.D. at another institution. For the last 20 years he works for the ONR in Washington D.C.

In short, I believe that motivation to learn comes from within and flourishes as one matures. It cannot be instilled if it's not there and the existence of grades is not an impediment to it. The impediment is the attitude that one has towards grades and that is what has to change instead of sunmmarily getting rid of grades. I think that is in agreement with what Alfie Kohn is saying.

 

[vela]

It has been a disaster. Students were not held accountable for the material and so multi-semester sequences (Physics I and II, for example) resulted in those students being unusually poorly prepared for the following course. Higher ed faculty are going to be dealing with pandemic-related K-12 learning deficiencies for years, and I'm not sure how that is going to play out. Ungrading will just exacerbate this problem because it removes a motivating factor for students to push themselves.

One of the complications in implementing ungrading is that students have to undo years of conditioning where they've learned the grade, not learning, is the ultimate goal. To successfully implement the ideas of ungrading requires that students buy into what you're trying to do. It doesn't sound like that happened at your institution.

It also requires instructors to abandon the idea of grades as well, and that's very difficult to do because most of us grew up in a system where grades permeate everything. Just look at this thread. Your response was the first one that really addressed my question. Everyone else talked about how they graded.

 

[mathwonk]

@vela; In one case what was asked was to make a presentation of some portion of the textbook. Maybe in some cases it was merely to attend class, stay awake and respond to my questions. These were classes of maybe 8 or ten students all of whom I knew well.

 

[symbolipoint]

Regarding the comments about grading on a curve, I don't do that- but not for the reasons given on this thread. I don't grade on a curve because then, ...,, some of my students MUST fail regardless how hard they work and try.

Why must they? How high was the class'es lowest score? Did your sense of judgement tell you that some students who learned enough still were part of the curve's lower end, and because of that place in the data, they need to have grade F or D?

 

[Dr Transport]

it was like pulling teeth to get many of my Asian students to talk to people outside of their peer group, and most would never ask me for help if they were stuck

Really, I had the exact opposite experience with the only Asian student in my class last fall. She wouldn't write anything down on a test basically without checking with me to see if it was correct. I finally had to send her away because she would write down an equation then ask if it was correct, then come back with the next line and ask etc... She told me she was afraid her parents would not pay for school if she didn't get almost perfect scores in all her classes.

 

[symbolipoint]

(I did not yet watch that video so this comment might be off-point.)

I can see at least one kind of problem with using pass/not pass, or credit/no credit instead of giving a letter grade.

If some group, institution, or organization has a requirement "Must earn at least an A in the course in order to enter the program", then the school where the student is earning the course credit cannot show the letter grade in documented form; student can only show, "earned credit" or "pass". The administration of the program the student hopes to enter expects to know the actual lettter grade. "Pass" might mean, C, or B, or A. "Pass does not limit itself just to "A".

 

[symbolipoint]

I just started watching that video. I am only about 15 seconds into it, and already I firmly disagree with the first research claim; that students become disinterested under the conditions to earn a letter grade. Do I/we really need to see the rest of the video?

 

[Andy Resnick]

Why must they? How high was the class'es lowest score? Did your sense of judgement tell you that some students who learned enough still were part of the curve's lower end, and because of that place in the data, they need to have grade F or D?

By defintion, 'grading on a curve' means that the distribution of grades is that of a normal distribution; the same percentage of students given a grade of 'A' is the same percentage of students given a grade of 'F'. Otherwise, it's not grading on a curve but some other engineered distribution.

 

[symbolipoint]

By defintion, 'grading on a curve' means that the distribution of grades is that of a normal distribution; the same percentage of students given a grade of 'A' is the same percentage of students given a grade of 'F'. Otherwise, it's not grading on a curve but some other engineered distribution.

That is helpful and now I really do not like that strict kind of "grading on the curve". This must be why many schools and institutions use a predetermined strict grading scale to assign letter grades of A, B, C, D, F. This way, the teacher assesses each test or quiz item the way he sees fit, and determines what the score is for the test or quiz, and likewise for other work assigned to the class members. At the end, the teach has maybe, percentage scores for all the students. I would imagine justifiably, any students in such a system who received a score like maybe 85%, are likely not getting D or F. But could be a C, or could be a B.

This "ungrading" business, I do not like. The only exception is for if the student wishes at the beginning or some such, to do a course for credit without earning a grade at the end. (Like at a c.c., student might visit the admissions and records office and switch his grade option for a course from Letter grade, to credit/no credit as long as the choice is allowed both for the course and the institution's official time limit.)

 

[Andy Resnick]

If my interpretation is correct, then one can argue that it shows that grades serve as an incentive to do better and, simultaneously, as a ustification to slack off. [...]

Grading can serve this important function, for sure.

 

[Andy Resnick]

This "ungrading" business, I do not like. The only exception is for if the student wishes at the beginning or some such, to do a course for credit without earning a grade at the end. (Like at a c.c., student might visit the admissions and records office and switch his grade option for a course from Letter grade, to credit/no credit as long as the choice is allowed both for the course and the institution's official time limit.)

That's an excellent example! We have an academic program targeted towards (ahem...) "older adults" called 'Project60" that allows participants to audit (the technical term for enrolling in a course w/o academic credit) any class they like.