During the last semester my supervisor Prof. X held a seminar for bachelor and master students (mostly computer scientists). According to the study regulations the students have to

  • give scientific presentations (based on journal or conference papers) and
  • hand in seminar papers on the topic of their presentation (i.e. their papers should cover the results of the original paper, lay out details of proofs, add additional explanations or examples, etc.).

The final grade consists of 2/3*presentation + 1/3*paper.

Prologue. At the beginning of the semester Prof. X's research assistants compiled a list of interesting research papers. At the first seminar meeting the students were able to choose the paper they would like to present to the rest of the group at the end of the semester. Furthermore Prof. X made it very clear that understanding the topic is just a portion of presenting and writing down scientific ideas. Every student was assigned to a seminar supervisor to whom they could talk when they ran into problems. Half-way through the semester there was an obligatory meeting with the supervisor and a deadline for handing in a draft of the presentation slides. At the end of the semester we organized a little "conference" and the students presented "their" papers. Afterwards, they received written feedback on their presentations from all participants (students, research assistents, and Prof. X). Six weeks later they had to hand in their seminar papers.

Problem. Unfortunately the quality of many seminar papers is relatively poor; even the papers handed in by students who had understood their topic "quite well" and had given good presentations are surprisingly different from what we expected.

Main issues:

  1. Even though they were allowed to use their native language, many papers had bad spelling (obviously no spell checker was used) and bad grammar.

  2. In some cases it was impossible to understand the basic ideas if one had not already been familiar with the subject.

  3. Imprecise language and almost no sources cited, e.g. "Algorithm Y is rather efficient in comparison with other algorithms." [citation needed].

  4. Some students only cited a single source (= "their" research paper).

Questions. How should we address these issues? Of course, we are going to give some feedback on their seminar papers. I am worried about the next seminar. Reading all these papers was (mentally) exhausting. Should we require that seminar papers must be handed in first? How could we install an iterative feedback process?

I am looking forward to your ideas and experiences.

  • 8
    For most people, it can take years to understand how to write decent papers: it seems to me that you want your students to accelerate this understanding in one semester. Commented Mar 24, 2018 at 13:36
  • I am aware of that and I don't expect a student seminar paper to meet the same standards as an accepted high-rank conference paper. I know that writing down scientific ideas such that people (who are still unfamiliar with the subject) can read and understand them. Especially this is hard for bachelor students writing their first seminar paper. But the issues 1, 2, and 4 (listed above) should not occur on a large scale.
    – diplo
    Commented Mar 25, 2018 at 9:13
  • Why not? Look at Stack Exchange questions across different sites: How many poorly structured, poorly written and unclear questions are there? How many people doesn't cite the sources of images, diagrams etc.? This happens also to native speakers and to graduate students. And a question is much shorter than a paper. For instance, for what concerns point 4, I had no concept of citation until I had started my PhD. This is the reason I stopped asking my students to write lab reports: they are generally awful and I don't have enough time and energy to improve a hundred lab reports. Commented Mar 25, 2018 at 11:06
  • The seminar papers are mandatory according to the general study regulations of our faculty. Hence, we must not skip this part of a seminar. To be honest, I was hoping that this thread would attract more attention.
    – diplo
    Commented Apr 1, 2018 at 11:11
  • Did you require draft papers from the students and give the students suggested corrections? Student papers will improve if you require this, particularly if you go through multiple cycles. It's very labor intensive though. Commented Jan 20, 2022 at 0:33

7 Answers 7


This sounds exactly like the situation we are experiencing at our university as well. We have improved the situation by forcing the students to write papers and similar texts every semester (starting from semester 1), in different settings, and we're giving detailed feedback afterwards. In our first and second semesters we are using this setting for one course:

  • Students can select one topic and prepare it in a group of three (the idea is to mix people so they can give feedback within their group). We emphasize that everyone will receive a grade based on personal efforts and success, and there is no group bonus / malus.
  • They will present the topic in a group of three (20 min presentation + 10 min discussion)
  • We are giving a 3h presentation on scientific writing.
  • They have to give their paper outline to their supervisor after about 3 weeks. The quality of feedback depends on the effort put into the structure. If they hand in "Introduction, Main Section, Conclusion", they will receive a "correct, but good luck" ;-).
  • They shall hand in a first version after about 50% of the course. This version will be discussed in a personal feedback session of one hour by a scientific journalist
  • At ~3/4, they have to submit their "pre-final" version to a conference management system.
  • Everyone is now switching roles and becomes a reviewer. Every reviewer has to review two papers. The quality of reviews is part of the final grade.
  • After the reviews. the groups have one more week to finish the final paper submission.

As said, we are following this highly intensive approach in semesters 1 & 2. In the subsequent semesters, we are having other settings, but we force them to write every semester. It seems to bear some fruits since the quality improves over time, but it's clear that we can't fully repair what went wrong during school...

  • Thanks for the answer. How does the peer review work out? This is an idea that we discussed internally, but we are not sure whether there is a constructive outcome. Do students see their peers' papers and think "Wow, this is so much better than mine. I will improve my own paper!" or are they not able to see that because if they could their own papers were better?
    – diplo
    Commented Apr 4, 2018 at 19:13
  • 3
    Quite well! Many students are criticizing things they did wrong by themselves and sometimes even learn from this 😉. For me it's a great help as well since I enforce plagiarism check in this step (without consequences for the authors at this stage). The feedback is quite helpful in general.
    – OBu
    Commented Apr 4, 2018 at 19:18
  • 1
    That sounds interesting. Based on my experience, whenever students have to give feedback to presentations, they tend to praise themselves (i.e. they defend themselves against the enemy aka the grader :-)). Hence, I would suggest not to grade the feedback itself. However, how can useful feedback be encouraged if there is no 'pressure'? Ideally students should give great feedback because because they need helpful comments on their slides and papers, too. On the other hand there are many reasons not to put too much effort into other students' papers instead of theirs.
    – diplo
    Commented Apr 5, 2018 at 9:42
  • 1
    We are using a nice trick: The review is part of the final grade. The grade depends on the quality and appropriateness of the review. This means, if they are way too positive, they will get a worse grade for their review. And I do not take the reviews into account when I'm, finally going through the submissions. So honesty pays of a) because they will get a better grade and b) their friends can improve their paper.
    – OBu
    Commented Apr 5, 2018 at 22:21

Here are some thoughts on the four points you mentioned:

Point 1. From my personal experience general spelling and grammar skills seem to reduce, but are most of the time a matter of "I don't care enough" or probably "I did not have the time to check".

As you already did a mandatory meeting with them half-semester for the presentation you may additionally request that they have e.g. about a page of their paper written with citations, etc. Then you can already point out some details ("there are many spelling errors, please use a spell checker" or "you do not give any references, where does statement X come from?").

Points 2./3./4. If the students have never written scientific paper (or something similar) before it will be hard. I second the answer by OBu in the sense that repetition and practice is key. Additionally, the paper they have read might be the first paper so they might not know how to cite and that non-trivial statements have to be either proved or cited.

The idea of everyone having a seminal supervisor they can talk to is great but it assumes that people will actually talk to them. Many might think that their problem (e.g. how to cite) is too trivial to bother the supervisor.

In general the things you can do are:

  • In the beginning give a lecture/talk about scientific working and writing. Show how to cite properly, that citing is necessary, how to find additional resources etc.

  • Make additional rounds of reviews, be it either between students themselves or by having the students hand it non-final versions of their paper to the supervisor and getting feedback on them. This also helps with the possibility of people just starting to write the paper shortly before the deadline.

  • Thanks for the answer. Having a designated supervisor and not talking to them seems like a key issue here. Setting up an iterative framework of meetings, feedback, beta versions of slides and papers might be a good idea.
    – diplo
    Commented Apr 5, 2018 at 9:50

Firstly, I think courses such as yours are a really useful opportunity for graduate students to practice academic writing and presenting. I did one with a similar marking structure during my masters degree and I found it a valuable experience. Here are some further thoughts on the points you are concerned about.

On point 1:

  1. It might be best for students to write in the language of the paper they have read. This would be helpful in situations where for example, field-specific words do not translate naturally into the native tongue and could instead be transferred from the paper. Of course, this might also worsen the quality of the written output.
  2. In my own experience, the act of preparing a lengthy presentation on a research paper for the first time was stressful and time consuming (especially as it was slightly outside of my field of experience). I used my seminal paper as a means to document my understanding as I went along and it was never my primary focus. If you want students to focus more on the written component, you could shift the grade break-down to put more emphasis on this part which will probably naturally lead to higher quality (at the expense of the oral component?).

On point 2:

  1. This is a hard skill to learn, especially if the research paper is not closely related to an area the student is already familiar with. Informally talking about it to their peers might be a good way to realize when they are not explaining themselves clearly. This could be done in small groups towards the end of the course, with the TA(s) moving around and providing advice.

On points 1-4:

  1. To some extent, this is probably achievable by having a very clear rubric for assessing the written work that stresses the importance of clear work with proper citations and use of multiple resources.

  2. Another option would be to have a short exercise at the beginning of the course (counting maybe 5%) where the students are given 2-3 short passages and asked to provide a summary, complete with citations. This could be marked fairly strictly, providing an opportunity to highlight issues.

  3. A third option would be to do something like a "journal club" where students distribute their paper to their peers before their presentation, and one component of the presentation involves a discussion of the work where issues are highlighted. The paper could then be revised before the final submission, allowing students to concentrate on and implement corrections that they have heard over the course of the presentations (although depending on the academic culture, this might be viewed as too embarrassing for the students).

  • Since these courses (seminars) are mandatory, we can't take credit for that. ;-) I agree that writing down in order to increase understanding is very helpful. If I cannot verbalize subject XYZ, I have not really understood XYZ yet. However, as you point out, this is time consuming.
    – diplo
    Commented Apr 5, 2018 at 10:04

It sounds like the oral presentations weren't so much the problem as the written work. Do you want to separate out the writing goal and work on that in a separate course? Or do you want to hold onto both goals, but increase the support given to the students on the writing prong?

If you choose the first option, you could offer something like a Technical Writing class, with a few group sessions and some one-on-one Teaching Assistant (TA) support, or you could ask the students to work with your campus Writing Center (if there is one).

If you choose to continue combining writing and oral presentations in one course, you could have them hand in the written write-up approximately halfway through the semester (but choose your date carefully, looking at the exam schedule), and do the presentations later in the semester.

Helping students become better writers is an iterative process, so plan accordingly and spread things out. You could have multiple due dates, the first for the outline (this will produce better writing and will help students get in the habit of starting with an outline), one for the content, one for the bibliography, and then a due date for the final submission.

Regarding the mandatory meeting with the professor -- consider switching this to a meeting with a TA, and consider requiring two meetings rather than one. (With just one meeting, both parties might be more likely to see it as a pro forma meeting.) I would put the first obligatory meeting pretty early in the semester -- but after you've had at least one group session.

Examples of topics to cover in the group session:

  • How having an outline helps.

  • How to install a spell checker for a non-English language.

  • How to add custom terms to the spell checker dictionary.

  • Common examples of plagiarism and how to avoid it.

  • One or two examples of famous people who fell into the plagiarism trap through sloppy scholarship.

  • Walk them through an example of an "A" project and an example of an unsatisfactory project.

  • How many references should they aim for? How do you decide what to include?

  • Later in the semester you can revisit some of these topics, once they've starting handing in some work; and also cover bibliography formatting and tools.

Overall, think in terms of supporting the development of better writers. Sink or swim assignments that get a judgment (grade) might make sense for most homework and projects, but not for this.

It might be a good idea to spread the presentations out over several sessions, for example four sessions, every other Friday afternoon, with cookies.

  • Unfortunately the study regulations are very strict: The seminar must include a presentation on a scientific state-of-the-art result to a student audience as well as a seminar paper. We are just free to offer additional help or to implement an internal seminar schedule with several deadlines, feedback sessions etc. You make a good point in suggesting that we go through examples of good and bad scientific writing and presenting. I think we can easily implement that.
    – diplo
    Commented Apr 5, 2018 at 10:15
  • @diplo - Not sure why you say "unfortunately." Most of my answer is focused on the combo course design (presentations AND written submissions). Commented Apr 5, 2018 at 12:11
  • I was referring to the first paragraph of your answer. We have no choice but to stick to the presentation+paper course design because of the study regulations that hold for all seminars in our bachelor and master programs.
    – diplo
    Commented Apr 5, 2018 at 12:22
  • @diplo - That's okay, you can make the course work either way. The main theme of my suggestions has to do with making the writing process more iterative and supported so that the students' writing skills improve. Commented Apr 5, 2018 at 12:25
  • I guess we are on the same page with that ;-). I just wanted to express that we cannot change the main goals of a seminar or drop the grading of the final papers. However we are free to implement extra tutorials on scientific writing, feedback sessions etc. (see your helpful comments) within the existing scheme.
    – diplo
    Commented Apr 5, 2018 at 12:35

Some of the following thoughts are already in the other answers, but i'd like to give a coherent answer, so please excuse any repetition. I've worked as a TA for several seminars which mostly were successful (and attended both good and less successful ones as a student), so here are some thoughts:

  • At my university (in Germany), there are normally two kinds of seminars. There are the normal seminars for higher level students, however before they get there, usually in the second or third semester, all students are required to attend a "proseminar" first. Outwardly, this looks similar to a seminar, however instead of focusing on current research topics, the talks are usually about something which is on the level of first year lectures but is normally not covered due to time constraints. (As an example back from when I studied computer science, I gave a pro-seminar talk on Petri-nets, which were not covered in the lecture on automata theory. Other talks in the same proseminar, were about different kind of state machines.)

    Inwardly, the idea is on one hand to allow a greater focus on the presentation and the written paper, as the topics are easier and there are normally good textbook sources, but on the other hand without completely boring the students, as they still learn something new. (Compare this to a course on "how to give a presentation", which most students will hate or willingly ignore, since many, like me back then, falsely think they know better anyway. This also is a good reason not to tell them that the topics are not the actual focus.)

    However something like this might be impossible to implement without reworking the curriculum.

  • Ideally you should know your students. This works best, if the seminar is directly subsequent to a lecture or if you have multiple seminars and "repeat customers". This way I often know, which students need strict deadlines and close supervision and which students can be allowed to roam free. Related to this, while there should be some choice in the topics, normally not all topics are created equal. Some are simply more complicated or have badly written sources, while others are mostly straightforward. Also not all students will have had the same lectures or even the same minor subject, so their previous knowledge of some of the topics may differ. A good way to handle things, is to present the topics to them and then allow them to give you a list of preferences (for example in the form of a doodle). This way they mostly get topics they are interested in, however you are still able to prevent people from taking on more than they can handle and you do not have to tell them that directly, as officially this is to prevent collisions.

  • Concerning the feedback, while peer review has its uses, it requires the peers to know what is expected. This works well for the actual talks, as the other students will notice if the explanations are incomprehensible or the slides are convoluted. However it works less well for the written papers or in order to gauge the correct understanding of the topic. This should be done by you and your colleagues instead. Ideally each student should have a direct supervisor assigned to them, who will be the main source of feedback and while in the end the professor has the final say, he will probably also be the one proposing the final grade on the written paper. This allows for more detailed feedback than if everybody has to read everything. You should however talk with the other supervisors regularly (for example at lunch) about what your students did and how you handled it, in order to have some consistency.

  • I always found it helpful to meet with each student one on one a few times, not only when they have questions, but also at certain milestones, where things are expected from them. An example plan would look something like:

    6-8 weeks before the talk, after they have read their sources: Go through the sources with them. Let them explain the details, check if they understood everything correctly and help them with their questions, ideally also pointing them to additional sources. Ask them, on which parts they want to focus their talk, tell them what should definitely be in there and which parts they should skip because of time constraints and instead cover in the written version.

    1-2 weeks before the talk, after they have prepared a first version of the talk: Go through their slides with them, try to correct problems and to estimate if the length of their talk is right. Ideally you should tell them to time a test-run themselves beforehand.

    After you looked at the first draft of the paper: Go over the corrections with them. For the first draft I will usually purposefully waste a lot of red ink and mark everything I see, from factual errors and missing citations down to typos, bad formatting and formulations which I simply do not like and personally would change. At this point I normally will tell them that they should not worry to much, since this draft will not count for their grade, but that I expect them to correct all those and similar errors for the final version. Remember to keep a copy of the corrections.

    After the final version is handed it: Give some final feedback. This is usually just a short meeting as most of the things will have been said beforehand anyway.

    This is very labour-extensive, which is another reason to split the workload among multiple people. I usually spend around an hour per meeting and at least an hour on the corrections of the first draft. On the other hand, reading the final version is normally just a cursory check if they heeded the advice. However spending so much time with a few students also allows you to get to know them a bit better, which tells you if they will need additional meetings or even to give a test talk before the actual one. The general idea is always to correct problems before they occur.

  • If the topics allow this, it can be helpful to group the students into pairs covering adjacent topics or two halves of the same. Let them attend the meetings together and possibly even write a joint paper. This way they know enough about what the other is doing to catch problems and they have a vested interest to do so. It is not helpful to force students to work together against their will, however often enough, some of them will already know each other. A good way is to explicitly announce some of the topics as group topics and ask the students to state if they want to work together when asking them for preferences.


I am currently undertaking a masters year project in mathematics which sounds similar in structure to this. In semester one I had a talk to give at the end of the semester. At the end of semester two I will have to hand in my final "report" (approximately 40 pages) and give a presentation on my project.

For comparison, I have been having regular meetings with my project supervisor, where the main topic of discussion has been the report which I am writing. Here are my thoughts on your main issues:

  1. With regard to the grammar issues, I think there are two separate things. Basic spelling errors can be fixed easily, as you say with a spell checker. Bad grammar is fixed by proof reading, but this is something which I can guess that the students are inexperienced with. Proofreading my own work is a skill which I have developed vastly during the time writing my project. I did not realize how much time it takes, nor how easy it is for mistakes to slip through the net. I suspect the students put a small amount of effort into proof reading but were not aware of how much time and care is needed to do this effectively. How you help them fix this problem is up to you. The problem was fixed for me by my supervisor relentlessly highlighting the mistakes which slipped through the net.
  2. This must come down to quality of writing. What would be helpful for me in this situation (if I were one of the students) would be to have a "guide" piece of work to compare with. Particularly in the case of lack of explanation of basic ideas, seeing it done properly elsewhere would probably be effective. I can imagine that part of the problem is that the students are worried about "dumbing down" their work, and just thought that there were various things that didn't need explaining. It might be worth reminding them that they shouldn't assume any knowledge on the part of their audience.
  3. Again, this sounds like something for which it would be useful to have a comparison. Perhaps in this case you could provide the students with a model "bad" example, and a model "good" example. That way, they can see which is closer to what they're doing. Giving examples (like the one which you have given us) of imprecise language and why it's not appropriate would probably be helpful.
  4. This sounds like inexperience on the part of the students. It is also (to my mind) indicative of a lack of wider research on the part of the students. Is this something which you have told them is important? Students are lazy.

I think generally, an iterative feedback process is extremely helpful, and would fix many of the above problems. That is because that has been effective in my case, and a lot of the mistakes which you are pointing out resonate well with me (particularly points 1, 2 and 3).


In my experience the problems you face are very common and I believe that it is almost impossible for the students to learn all the necessary skills within just one course. In order to address these problems, together with colleagues of mine, we tried to introduce parts of these skills in a variety of courses starting with the introductory courses.

In our study programme there are multiple exercise classes where the students are supposed to solve problems and to present their solutions. Traditionally their presentations were graded only based on the quality of the solutions themselves. For these courses we introduced a grading scheme for these presentations which take the quality of the presentation into account. This grading scheme contained different parts which address different aspects of the presentation, e.g. whether the problem was explained or whether the necessary background was presented.

Since these students also had to hand in some of their solutions we introduced a similar grading scheme for these solutions which again took into account the quality of the exposition.

In more advanced courses we combined these grading scheme with some kind of peer review, i.e. the students had to grade some of the "papers" and some of the presentations.

In our experience, these exercises improved both the quality of the students' writing and of their presentations. It also seems that the existence of an "official" list of criteria for a good presentation or a good written text already improved the quality.

You must log in to answer this question.

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