How to make a good short report

The end of May and the beginning of June is the traditional time for the protection of diplomas in universities. In addition to the final work, students need to prepare a short (for 7-10 minutes) presentation before the examination committee, as well as make a presentation. And often for a student to prepare a normal report / presentation is almost more difficult thing than writing a diploma itself. Moreover, it often turns out that the more informative and interesting the work, the more difficult it is to make a high-quality performance on it. In this article, we offer our view on how to solve this problem.



We in St. Petersburg HSE traditionally spend April-May to work with graduate students on their defense speeches (and the previous six months to hear full reports). And we have somehow come to terms with the fact that, despite the very high level of our students, few of them manage to go through a rehearsal the first time. As a rule, we have to listen to reports 2-3 times, edit presentations, add / modify / correct the presentation of material before declaring the presentation suitable and releasing the student for protection. And of course, from year to year we have to explain to students the same basic principles of building a report, organizing a presentation, etc. This article is some attempt to present in writing all the main points that we traditionally explain to students who come out on defense.And we will be glad if the considerations that we present here turn out to be useful not only for our students, but also for all other students - Habr readers coming this year to defend their diplomas, as well as all those who have to prepare short presentations for work according to the results of their current work.

Note: further illustrations are taken from the presentation for the thesis of the first year student of our master's program “ Data Programming and Analysis ”, graduate of the bachelor's program “ Applied Mathematics and Computer Science ” Egor Bogomolov.

1.First of all, the format of a 7-10 minute presentation significantly limits the number of slides in a presentation. Experience suggests that it should have no more than 12-15 slides. Why? If we take a 7-minute report and divide 7 by the number of slides, then on average we get about 30 seconds per slide. 30 seconds is a fairly short period of time, it is hard to say something meaningful during this time. If the number of slides becomes larger, then either the report turns into a machine-gun queue and no one has time to understand anything, or the speaker is guaranteed not to fit into the time allotted for him. Based on this, let's fix the number of slides (12–15) and understand what should be on them.



2.With the first slide, everything is more or less clear: the title of the report, author, scientific adviser, organization is indicated on it. It is also advisable to mark the city and the year. The only subtle point: I want the information placed on the first slide to be properly formatted. This slide sets the style of the whole presentation: the style and size of the pin, the colors that will be used later. For example, we recommend using blue and white - the corporate colors of the Tower.



I’ll tell you about the style of the remaining slides. On each of them it is desirable to do both the header and footer. It is desirable to indicate the name of the presentation block in the upper one (let's say about the blocks below), in the lower one - the speaker’s last name, the short name of the report, and, importantly , the number of the slide in x / y format , where xIs the number of the current slide, y is the total number of slides. Why is this done? Listeners, as a rule, want to understand at what point in the report they are and how many more slides are ahead. If this information is not available, then either the impression is created that the report can last forever, or that it can end at any time. This is annoying.

3. Now let's move on to the content of the slides. The entire report can be conditionally divided into two parts: the introductory part (before formulating the goals and objectives of the work) and the substantive part (after the goals and objectives). Important : in a good report, the introductory part takes about half (!)the whole presentation, and here's why. If the speaker does not make a normal introduction, does not get everyone present to understand the topic of the work, the statement of the problem, etc., then no one will listen to the substantive part - the audience will simply not understand what the talk is about. Therefore, 50% success is a well-constructed introductory part.



Important : from our point of view, the introductory part must consist of three blocks. First block- statement of the problem, in it the speaker tells what the task is, what subject area it belongs to, introduces basic concepts and definitions, indicates why the solution to this problem is important. And this block, with rare exceptions, do not forget to include everything in your report. However, not everyone, alas, knows how to tell her correctly. The speaker’s typical misconception is that almost everyone knows about this task, except for small details. And instead of explaining on the fingers what his task is, the speaker begins to freely use the terms, definitions and facts, known, as a rule, only to him and a narrow group of his colleagues, without explaining them in any way.

I can say a rather seditious thing, but the statement of the problem should to some extent be similar to the answer to parents / friends / acquaintances in the bar to the question “What are you doing at your university / work?” And it is unlikely that sitting in a bar or in the kitchen with your parents, you will pour in unfamiliar terms and build yourself into a cool specialist. Most likely you’ll try to explain with your fingers at least the subject area in which you conduct research. And as a rule, while sitting in a bar / in the kitchen, you manage to do this, but during the report - for some reason, no. You must understand: the simpler, clearer and more understandable you are about your subject area, the more people will listen to your report to the end.



Let's move on to the second block- A review of the results available in this area. Most of the speakers have trouble with him. For some reason, it is believed that it can be safely thrown away, especially in the case of short reports. Meanwhile, this is a colossal fallacy; this block is categorically necessary. If it is not in the report, then this usually means two options.

The first option - the speaker simply does not understand that, in addition to him, there are hundreds, if not thousands, of researchers working and receiving results in this field. As a result, it is highly likely that everything that the speaker is about to talk about has already been done by someone somewhere a long time ago, and he is just reinventing the wheel.

Second option- that the task is so uninteresting and unnecessary to anyone that no one is doing it. Therefore, the relevance of this task is close to zero, and listening to this work also makes no sense. These, of course, are extreme cases, there are intermediate ones. For example, a couple of references to 1994 works. And if that's all - then again trouble, see the two options discussed above.

Dear colleagues, without reviewing the results of other researchers, all your work has absolutely no meaning! In it, you must convince listeners that you are aware of all the latest research in this field, fully own the material, know who, what, and when did over the past 4–5 years . If this is not the case, you can not listen to the report.



It would seem that in a review of the literature you can finish the introductory part and proceed to the formulation of the goals and objectives of the study. Actually, most inexperienced speakers try to do so. And this is a very gross mistake: in any presentation, there must be one more extremely important block, namely, a block under the code name “Big BUT”. In this block you should say something like this: yes, a lot has already been done in my area (see the mandatory 1st part) (see the mandatory 2nd part), but: there are such and such drawbacks in such and such a work, in such this-and-and-such has not been done to this work, but this-and-and-such remains incomprehensible from this work. And in my work, I am going to eliminate these points / to complete these unfinished tasks / to offer my own, more successful solutions. And only after that you can move on to the central part of the report - the formulation of the goals and objectives of the work .



4. The goal of any work should be one. Multiple goals are not allowed. If there are several of them, it is, rather, in the wrong place and in the wrong form that the formulated objectives of the study. purpose- this is something that you integrally want to achieve with your work: create a software product, assemble a dataset, build a model, conduct an analysis, etc. Think about what you want to achieve with your work. As a rule, this will be your goal.

Important: in very many works, a rather significant mistake is made in this place - the goal is formulated too generally, for example, “creating a spam filtering system” or “detecting bots on social networks”. Colleagues, these are tasks that for several years have been solved by huge IT companies such as Google or Facebook. And it will be very funny if it suddenly turns out that it is in your work that this global task will be finally solved. With a very low probability, listeners will believe in such an optimistic scenario. Most likely, a similarly broadly formulated goal is still, as a rule, a sign of inexperience and / or incompetence. Please do not make this kind of mistake - state your goal very specifically.

Let's move on to the tasks. Tasks- This is, in essence, a short plan of your work, a brief description of how you are going to achieve the goal set above. For example, “you need to create a model”: 1) create such and such a tool; 2) to collect such and such a dataset; 3) build a prototype model; 4) test this prototype on such and such data. Often, in order to formulate tasks correctly, you just need to look at your work once more, break it into 3-4 large blocks and somehow call these blocks. Actually, these will be your tasks.

Important : there should not be any new terms in the goal or tasks. If they suddenly appear, this means that you have not completed the statement of the problem, have not fully explained all the terms, tools, mechanisms that are usually used to solve such problems, etc. Further, there should not be many tasks. As a rule, these are 3-4 global tasks. Do not crush them, do not do 7–8 tasks. Finally, tasks should be sufficiently short and concise . Ideally, each of these tasks should be placed in the footer on subsequent slides.



5.We now turn to the second, substantive part of the report. Its construction is very simple. You have about 7 slides left in stock. You should leave the last slide for conclusions, 6 slides remain. These slides are divided into blocks according to the number of tasks. And for each task in your short report you leave 1-2 slides. Alas, this is the maximum that you can afford because of the time limit. What to bring there is already a rather specific question, everyone should solve this problem for himself. It is important here to choose truly important points in each of the tasks and try to briefly talk about them. There are not enough general recipes for this part, and experience is very important here: yours, your supervisor and your colleagues.





Often in this place, students ask the question: well, well, I somehow somehow schematically, on my fingers, have time to show something there on 1-2 slides; but there are a lot of important details, dozens of important details, and not a single professional listening to the report will be able to evaluate the work without this additional information. What to do in this case?

Here you should understand the following. As a rule, there are only 2-3 people among the audience who really understand what your report is about. Most likely, they would really be interested in learning about the details of the study. For the rest, the details are not important, they need a general idea. Explain it on your fingers, and for narrow professionals, prepare from each block several additional slides outside the presentation. For example, your report has 15 slides. Leave the numbering as x / 15, where x is the number of the current slide. Behind the 15th slide, make additional 16-17th slides in the format x / 15, x = 16.17, and so on, to which information that could not fit in the main blocks can be taken out.



If suddenly (which is not a fact) after your report one of the listeners decides to ask you a specific question, you will open these slides and answer this question in detail. At the same time, answering the question, you will no longer be limited to those 7-10 minutes that were given to you for the report. Therefore, you can add that information that physically did not fit into the main report, but about which you so wanted to tell something.



6.Go to the last slide. On this slide, you state the conclusions or results of your research. In fact, here you just have to duplicate the tasks that were formulated in the middle of your report. However, the question arises: why then is this slide needed at all? Isn’t it easier to just show the tasks again and say that they are all completed? No, there are important differences. The fact is that, while formulating the objectives of the study, you still cannot appeal to the information that you will have in the content of your report. You have not yet disclosed with the help of which ideas, methods and approaches you have achieved your goal, how successful your approaches have been, how well testing has been done, etc. However, at the end of your report, the audience is already immersed in the topic, they know the terminology,the solutions you have proposed are known, and you can safely reveal the tasks formulated above in the conclusions and briefly describe how exactly you solved these problems.

Well, the cherry on the cake in this part of the work will be, for example, links to articles that you wrote as a result of the study, reports that you made at a particular conference, software products that began to use the code you wrote, etc. .

Small note: the final slide a separate "Thank you" is not necessary - this phrase perfectly calm one can say at the end of words. A slide like “Your questions?” Is also inappropriate. This phrase should be said by the person who presides at this meeting, do not take away his bread.

7. Well, in conclusion, a few words about what to do if you do not have a short, 7-10-minute, but a large, 30-40-minute report. Important:the entire structure described above categorically remains unchanged. There remain the blocks “statement of the problem”, “review of the available results”, the block “but”, the slide “goal and objectives”, the slide “conclusions”. What is changing? The number of slides is changing, primarily in the content part: now instead of 1-2 slides, you can finally insert 5-7 detailed slides for each task.

But this is not all: expanding the content part, you should not forget about the introductory part. It should also be expanded: it is necessary to tell in more detail about the statement of the problem, to give a more detailed overview of the available results. Remember that the introductory part remains at least, if not more, an important part of your report. Do it poorly - and no one will understand all the remaining information, or no one will listen.

That’s all in general. We hope that these notes will be useful to everyone who is preparing their reports, and especially to those students who are preparing presentations of their graduation works right now.

PS Here you can see diplomas and presentations of graduates of our bachelor's program “ Applied Mathematics and Computer Science ” for 2019. Of course, they are far from ideal, and they could (and should) be improved, but they can give an idea of ​​some final part of the great work on preparing our students for defense.