I'm an undergrad teaching assistant of a physic BSc student's lab. My job is reviewing the (usually 8-10 pages long) reports the students make, arriving in electronic format. During the review, I add short comments (1-2 sentences) to the document: things like "that's false because this and that", "this would've been better such and such", "you also could've check that and that", etc. The students then get their commented report back graded by the professor. Related to the course, I never meet the students in person.

The main purpose of these comments, among letting students know what they did wrong, is to help them improve their experimental skills. Observing and documenting everything reasonable, interpreting the results, thinking further, being consistent, etc. As such, there're usually a lot of comments on every student's report - usually 3-5/page - even on A-graded ones.

My concern is it's kind of hard to write even constructive comments of this kind without the possibility of sounding somewhat cantankerous. The relatively lot of "that"s bad, that's wrong/missing/false/etc" comments I make on small, but not unimportant errors set up a negative tone as is, and in this atmosphere even writing something like "that wasn't strictly in the task, but it's interesting to think about this and that" may come down as negative. I try to add as much positive comments as I can - but the simple "Good!", "Clever!", "I like the approach" kind of notes feel falsely in large amounts, especially when the student didn't really do anything outstanding, only did what he was told. Expanding them with something like "That's good work. You could also check that and that" also turns into education.

I'd like to avoid coming down too testy: it would kill the purpose, making students dismiss my comments as "nah, that guy finds error in everything". I want the students to feel that I actually want to help them, and I'm on their side. The professor - of course - told them the purpose of the reviews, but I know that when I did this same lab course (a year ago), there were still (silent) words like that about our teaching assistants.

How could I make my coments more encouraging, without sounding false?

  • 2
    My answer to How to find positive things in a code review? on Software Engineering SE has application here.
    – Wildcard
    Apr 3, 2018 at 23:50
  • 1
    You may be looking for the term 'teaching assistant' (TA), which is relatively standard phrasing.
    – E.P.
    Apr 4, 2018 at 8:13
  • I just wrote testy comments, but I was dealing with a batch of kids who couldn't wrap their minds around the difference between a mathematical proof and an example. -_- Apr 4, 2018 at 14:00
  • You are probably going to seem a bit testy to them in writing. That is okay. It's part of your job to show them where they are wrong. Just make sure that when you talk to them in person you are constructive and upbeat. Apr 4, 2018 at 17:32

2 Answers 2

  1. I can see you've given some thought to balancing the negative with some positive. That's good. Note, you don't need to remark on every single thing they did right.

    If you have a rubric, a list of elements you look for in each lab report, you may wish to put a check mark in the margin wherever you see that a required element is present.

    Sometimes a student doesn't manage to get much of anything right. In that case, it can be helpful to notice something positive about what they were trying to do, e.g. "Glad to see you caught this contradiction."

  2. Try expressing things as questions, rather than statements. Examples:

    Can you also check such-and-so? How does this fit in with etc.?

    One of my professors would write in my proofs, instead of "Does not follow" or "Not relevant,"

    How is this germane?

  3. As a student, with certain graders, I was sometimes unsure whether a particular comment was positive or negative or just food for thought. So try to avoid those misunderstandings, e.g. "Food for thought: etc." or "Enrichment idea: etc." or "Conclusion is correct; be careful to make clear the connection between…"

  4. Be tactful — meaning, your belief that they are trying hard, and can get better with your guidance, should come through. Example: "Watch out, need more digits here to avoid rounding error later on."

  5. I suggest you arrange with the professor to visit the class briefly at or near the beginning of the semester, and maybe at the midpoint again, to introduce yourself and show a friendly face. Convey to them that your job is to give them lots of specific feedback to help them improve their work. This can help humanize things.


Let them think you're testy or overly picky. They will still take your feedback on board if they care about doing a better job next time--either for the sake of their own skills or simply to get a better grade. And yes, they'll complain about you, but that's no reason to worry. Thinking back to my student days, the professors and graders that my classmates and I complained about most, were often the ones we learned the most from.

Regarding tone, you're right that a neutral tone can come across as harsh in writing, but it's something people can get used to from you. In fact, once they do get used to it, this way of communicating is refreshingly efficient.

At most, you could ask the professor if he thinks the tone of your comments is okay, but I suspect you don't need to change anything.

  • 8
    Agree -- I might add that one could distinguish between major errors and other comments. For example, the major errors could be red bold. Then the students will learn that only the red bold indicates a real problem, and everything else should be taken in a scholarly spirit. (Personally, I also like the idea of using a different color for very positive comments, but you want to limit the amount of time you spend color-coding).
    – cag51
    Apr 3, 2018 at 20:31
  • @cag51 Depends on the person for sure, but I've used the color-coding approach to great success. It helps when you're going over the papers to assign actual grades.
    – David Z
    Apr 4, 2018 at 4:27
  • @cag51 Thanks for the idea. After thinking for a while, however, I decided againist using colour code. Distinguishing major errors implies that the others are minor, and the students will be much less motivated to actually care with them. I'd rather like them to care with all, be it a simple spelling or a skipped task.
    – Neinstein
    Apr 4, 2018 at 7:17

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .