Do you allow students to submit "test corrections"? Meaning, do you allow them to take the questions they missed on a test, denote the correct answer in some way, and give them partial or full credit back if they then get the answer right?

If so, why? If not, why not?

I'm wondering, because I have a colleague who has ALWAYS allowed test corrections in her CS courses for half credit. I NEVER have. But this year, for some reason, I am getting a lot of lip from students like "What! No test corrections?! But Ms. Xxx always let us do test corrections!"

I believe allowing test corrections (which is essentially a retake) does a disservice to the student because it allows them to be lazy. They don't do it right the first time. They don't study (or don't study as much as they should have) and depend upon "corrections" to make up for their laziness.

My colleague believes allowing test corrections at least forces the student to "discover" correct answers to questions they may have never revisited otherwise, and that therefore they have another opportunity to learn.

Am I being too hard? What do you think?

  • 1
    $\begingroup$ Depends: Are you trying to facilitate learning, or measure ability. If the First then see Carol Dwekes work, and give grades of “Not passed yet” or “60% so far”. If the second, then do an entrance exam, but the grade in a box, and re-issue on graduation (you don't want to affect what you are measuring). $\endgroup$ – ctrl-alt-delor Mar 7 '18 at 19:08

Our district has a policy that any student that scores under a 70% must be allowed to retest. You take the retest score, average it with the original, and cap it at 70%.

Catch is that I teach all AP / Honors classes. Typically these aren't students that score a 60% and want to bring their grade up to passing. They're the kind of student that gets 88% and is unhappy they didn't score an A.

I go beyond that and let anyone who wants to do test corrections instead of a retest and they'll get half their points back. We have another AP teacher that does a square root curve if they turn in corrections.

To get credit they turn in a document with a screenshot of the question (tests are all in Canvas), explain why they picked the answer they did (sometimes "I guessed" is okay), what the right answer is, and most importantly why it's the right answer.

3 reasons why I do corrections.

1 - It lets me throw a few harder questions into a test. I start putting in AP style questions early in the year, which students may not be used to. As the year progresses more and more AP style questions get put in. If the longer questions go in, especially early in the year, the tests might be a little too difficult.

2 - I don't really care what they know Tuesday at 10:30. I care what they know when they leave my class at the end of the year. If they can come back after a test and prove they know the material, that's much better than just leaving it unknown.

3 - Everybody has a bad day. Maybe their dog died last night and they're not in a good mindset to take a test.

With all that though, I'd rather do retests. Still for half credit. The problem I'm facing with that is I'd essentially need at least one more full test for every unit. Otherwise they're retaking the same test. What I'd really like is to have enough questions banked that students could take a retest 2 or 3 times and I'd average everything together. But that would take hundreds of questions for each unit.


Yes, allow it, but control it a bit. The only reason I didn't permit it on final exams was the lack of time.

However, you don't have to give full credit to points gained on retakes. I used a rule of about 80-90 percent, which was enough to limit laziness, at least.

Remember that your job is not grading students, but educating them. If they get something wrong it gives them an incentive to revisit it immediately and get it right. The reinforcement of both looking and re answering can have a positive effect.

I'm not teaching anymore (retired) but would permit redoing work (including tests) even after I'd discussed the task in class.

If they don't get it right the first time, they need more reinforcement. Make it so. But the small penalty is enough to goad all but the laziest of the flock.

And recall that sometimes the best students get a question wrong because there is something ambiguous about the asking.

Like many things in education, you can try this for a while and evaluate the effect. Then make adjustments as needed.

  • $\begingroup$ "Remember that your job is not grading students, but educating them." + 1000 $\endgroup$ – Kevin Workman Mar 5 '18 at 23:55

I don't go in for the notion that students will be perfectly happy to laze around and get partial credit (50% is a very steep penalty!) In lieu of trying to get a genuinely good grade if they feel like they can.

I don't allow test corrections, instead I have a full system of progressively (slightly) harder make-ups which I have described here in some depth.

If the goal is to make sure that students learn the material that you are teaching, then the impulse to create some sort of additional pathway to learning after an assessment makes a lot of sense. And just after an assessment has gone badly is a perfect time to get strong, focused learning from students who, perhaps, thought they knew the material better than they really did until but a few moments ago. The iron is so hot at that one moment, it's really a pity if you don't strike it.

Why do I say this? Remember that strong emotions are one of the neurological keys to immediate and impactful learning. Thus, there may simply be no other opportunity like this to reach a student who, perhaps, is not entirely in love with the material, but simply wants to do well and get through. This is that fleeting moment when you can help them understand that idea that they didn't quite get the first time, because this is when they really care. They messed up, perhaps badly, and if they really focus right now, they might be able to partially fix this.

And maybe, just maybe, you'll get them through a tricky idea that will really help them later. Such as later in your same course, when they will really need to have already mastered this material if they are going to be to even have a good, fighting chance at the new material.

I've chosen for myself a very different structure than test corrections, and what I do works very well for me and my students. I might have to recalibrate if I taught a different population. But I suspect that whatever methodology works for you is less important than that you do something meaningful with those failed assessments, because they can be one of the most powerful tools you have, and you ultimately don't get too many of them.


First Two Years - Yes-ish, Juniors: Maybe, Seniors: No

From your profile I see you are teaching high-school.

Let me preface this by saying that I favor practice-tests over test-corrections.

Some points:

  • Test corrections are a security blanket for students who want to parlay a 70 into an 85. You will be popular for it.
  • Test corrections will likely never be used by those who need them most. Your worst performing students are often your laziest, and the kids with the lowest marks will likely not submit test corrections.
  • To follow the previous point, this is a great defense against parents offended by your inability to give their child higher grades. They could have easily gotten more points, but didn't bother submitting corrections.
  • Test corrections do not work for all types of tests: Multiple-Choice, Fill-in-the-blank, anything overly based on memorizing something specific. It does apply well when testing concepts and practices.

On Seniors:
If you are teaching seniors, then no, no test corrections.
When teaching seniors, you not only need to teach subject material, but also how to survive post-secondary education, which will mercilessly slaughter those who bank on second chances.

Explain to your seniors that you are preparing them for reality, and while they won't be getting murdered by a 20%/80% Midterm/Final combo strike, they'll still have to take their tests seriously.
However, I do recommend giving practice tests.

For younger students:
I don't see a problem with test-corrections. Of course, I would certainly only offer half-credit, at a maximum.

Further, I would require a write-up explaining why their first attempt was incorrect. If they want to earn some points back, they need to be able to teach the previously flawed concept back to me.

Practice Tests:
Do a practice test for half of a class period, then in the second half, grade it as a group.
Poll the class: What questions were commonly missed? What misconceptions were present? How "tricky" does the class rate each question? A class of highschoolers loves to comment on how difficult a problem is, or how sinister they believe their teacher is being.

Regarding forcing students from laziness vs allowing them to re-discover:
You are teaching high-school-students. This means you have a very mixed bag.

You absolutely have students who will go on to use none of this at all, you have students who just want to sleep all day and only work on their education later, and you have students who actually want to succeed.

In reality, if you gave devastatingly difficult tests with no test-corrections, they are still able to re-discover. The result of one test will not end their academic career. In university, yes, one test could set you back a full year, depending on how your prerequisites are structured. In high-school, no. Getting a bad grade on a test will not kill you, and, you will likely still be using the same concepts throughout the remainder of the year.

A test that you cannot re-submit for corrections is not a closed book.
It's still a learning opportunity. It still points to areas of improvement.

In CS, it is very rare that a concept will only be used once in a high-school level course, then discarded.

A student can certainly learn from their mistakes, to their benefit (using the same tested concepts on a future test or project), without receiving partial marks from a previous test.

Later in their academic career, a student will tend to value that "tough" teacher, who was harder to earn points from. I've heard: "The students who took Ms A's class actually know how to do this stuff... I got stuck with Ms B, and we all just slept through class and got perfect on everything..."

Test corrections is not for everyone. If your class is structured in a way that the grades fall where they should fall (instead of everyone getting roughly the same watered-down grade), then you have little motivation for allowing test-corrections beyond popularity. In fact, it could dramatically affect your class ecosystem.

If, however, you feel motivated to change some things. Consider allowing test-corrections with a low value of extra points awarded for considerable additional effort.
By doing this, you are rewarding aptitude and effort.


Some teachers feel it's cruel to treat 4th year classes differently from 1st or 2nd year classes, instead favoring consistency.

While I see the value in consistency, I place more value on evolution. However, there is certainly value in both styles.

If advancing expectations are outside of your teaching style, I'd recommend against introducing a crutch that requires weaning.

There are certainly other ways to allow a student to learn concepts they are deficient in. Ensure you have an avenue for students to learn from their mistakes, and ensure that avenue travels through your teaching style and class format.

  • $\begingroup$ Hmmm. We college/university teachers may not, in fact, be the monsters you seem to imply we are. I also think that students in their future will tend to value those who show human qualities rather than keeping strictly to rules that have no exceptions (robot qualities). Yes, I was seen as a very demanding teacher, but I was, at least after I learned to be, always kind and understanding. I don't, however, disagree with all you say here. $\endgroup$ – Buffy Mar 6 '18 at 17:11
  • $\begingroup$ @Buffy To the unprepared student, the harsh reality of most post-secondary credit systems will hit like a train. Students regularly fail and drop out of university courses taught by the most benevolent, most talented, least-monstrous professors possible because they will ill prepared, and didn't take their tests seriously. $\endgroup$ – Gorchestopher H Mar 6 '18 at 17:45
  • $\begingroup$ Perhaps, but treating them badly (IMO) now so that they will be prepared to be treated badly later doesn't seem like the best solution. There must be a better way. However, our chat room is a better place for this kind of discussion: chat.stackexchange.com/rooms/59174/the-classroom $\endgroup$ – Buffy Mar 6 '18 at 17:59
  • $\begingroup$ @Buffy I can continue in chat, but I'd like to state that I don't believe anything I'm suggesting, nor the OP's current test practices constitutes treating them badly. $\endgroup$ – Gorchestopher H Mar 6 '18 at 18:17
  • $\begingroup$ Thank you @GorchestopherH for your response. I agree with a lot of what you say. I try to walk the thin line of being kind to my students and developing relationships with them, while at the same time preparing them for the harsh realities of life in secondary education. I also do not think this constitutes treating them badly, nor do I think this in any way implies that college/university professors are monsters. I know you guys took this to the chat room, so if you have a response I'll be glad to join you there. $\endgroup$ – Java Jive Mar 7 '18 at 13:14

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.