How to write a thesis

Inhaltsverzeichnis

1 Basics for Success

The following factors make a good scientific work:

  • clear problem/objective
  • logical, structured layout accurate
  • handling of terms, plausibility comprehensible reasoning (through clean structuring, argumentation, references, objectivity, etc.)
  • content and formal accuracy
  • systematic approach and critical questioning of results
  • interesting presentation of facts (also through good illustrations, etc.)

back to table of contents

2 Formalities

2.1 Parts of a Student Thesis

It is best to use the templates (Word and LaTeX) provided by the department, especially to automatically generate the title page, declaration of independence, and all directories. The thesis must be written according to the following structure:

  • Title page with the following contents: name of the university/faculty/institute/department, name of the university professor, type of work, topic, your own name, date and place of birth, name of the supervisor, date of submission
  • Task description: Take over the text of the task description unchanged and in full
  • Declaration of independence
  • Summary/Abstract (half a page each): It must stand alone and provide an overview of the entire work (including results). Do not use or introduce abbreviations! No references to parts of the work!
  • Table of contents: Listing no deeper than level 3
  • Symbol/formula/abbreviation directory (optional): Used symbols and terms can be collected here in one place
  • The actual text of the work: The first chapter (Introduction) begins here and the page numbering. The chapters are numbered decimally (do not use more than 4 levels of structure, better 3)
  • Bibliography
  • List of Figures (optional)
  • List of Tables (optional)
  • Glossary (optional)
  • Appendix (optional, alphabetical numbering of chapters)

2.2 Structure of the Thesis

The IMRAD (Introduction, Methods, Results and Discussion) schema is a common standard. Proposed structuring:

Chapter 1: Introduction Motivate the problem in the application context, restate the task in your own words, and give an overview of the work.

Chapter 2: Related Works Show who has already dealt with the topic or related topics, what solutions have been described, and what the connection is to your own work.

Chapter 3: Fundamentals Introduction of mathematical, technical, algorithmic, or other basic knowledge necessary to understand the work.

Chapter 4ff.: Methodology and Implementation Main part of the work - first describe the concept, then the realization.

Chapter 4ff.+1: Results Objectively present the results and describe how exactly they were obtained. Draw attention to peculiarities.

Chapter 4ff.+2: Discussion Discuss the implemented solution based on the results. Understand and explain peculiarities. Based on this, work out the pros and cons (of the developed method). Under certain circumstances, the chapters "Results" and "Discussion" can be summarized in a single chapter.

Chapter 4ff.+3: Summary Brief summary and evaluation, results/solutions are condensed into a conclusion.

Chapter 4ff.+4: Outlook The outlook shows meaningful possibilities for further processing the material. Chapters "Summary" and "Outlook" can be summarized in a single chapter under certain circumstances.

 In Chapter 7↓ (Best Practices), you will learn more about the contents of each chapter.

back to table of contents

3 Formatting

It is best to use one of the chair templates. You can find these at: tu-dresden.de/ing/informatik/smt/cgv/studium/materialien

3.1 Use of Fonts

Body text in serif font creates good readability. Headings can look nice in sans-serif bold font. Recommended font sizes are:

  • Body text: 11pt
  • Headings (h1-h2-h3): 18pt — 14pt — 12pt

Font Styles

Emphasis can be achieved through italics, bold, ALL IN CAPITAL LETTERS, small caps, and font family. However, do not use several AT ONCE! Underlining is prohibited, bold and capital letters should be used sparingly! For source code, a monospace font is recommended (e.g., Courier New). Free variables and free function names should be italicized, whereas characters with fixed meanings should NOT be italicized - these are well-known functions (e.g., sin/cos, lim), constants (e.g., Euler's number e, the constant π, or user-selected constant symbols), and unit symbols (e.g., m/s, kHz). Mathematical variable names never consist of more than one letter! If more characters are needed for precision, they may appear in subscript (also not italicized). Examples of font formatting in formulas:

  • Incorrect: func(x) = πxmax*sin(2x)
  • Correct: f(x) = πxmax⋅sin(2x)         Note1

Italicized text can also be used to indicate (self-introduced) technical terms and foreign words and bold can be used to indicate keywords. Furthermore, italics are used when referring to titles of independent work (monographs, books).

Quotation Marks

Use quotation marks “” not to emphasize words, but exclusively to quote text passages or when referring to non-independent literature (articles from conference proceedings or journals, essays, book sections).

3.2 Page Layout

The work should be printed double-sided. Leave margins for notes. A good layout has an outer and inner margin of 2.5cm each, as well as a top margin of 3cm and a bottom margin of 2cm. The top margin leaves room for a 1cm high header, which bears the title of the current chapter and the current page number on the outside. The page numbers begin on the right (odd) page with the introduction. Chapter beginnings are always on a right (odd) page. If necessary, the preceding (left) page remains blank.

3.3 Recurring Elements

Figures

are to be labeled according to the schema Fig.␣<Chapter number>.<Sequential number>:␣<Title>. This refers to captions that are placed below the figure. Find a meaningful title. The figure must be self-explanatory along with its title. Pay attention to high quality, and prefer vector graphics. Also, ensure that each part of the image is large enough and that the captions are large enough for good readability.

Tables

are to be labeled according to the scheme Table␣<Chapter number>.<Sequential number>:␣<Title>. This refers to headings that are placed above the table. As with figures, a meaningful title is important.

Source codes

are to be labeled according to the schema Listing␣<Chapter number>.<Sequential  number>:␣<Title>,␣<Filename>. For short code passages, captions can be used; otherwise, use headings.

Footnotes

for annotations or translations. If the footnote refers to a word, the footnote mark immediately follows it. If it refers to a sentence, it is placed immediately after the period. Use footnotes sparingly. Consider how the content can be incorporated into the text in a meaningful way.

back to table of contents

4 Citations

All sources, including texts, images, surveys, links, etc., must be cited. The author and the source of the content (books, papers, slides, web pages, etc.) should be identified in the bibliography. The reader must have a complete overview of the sources used and their origin, especially for non-printed media. Permission from the author is not required.

4.1 Citation Style (with Square Brackets)

A citation is marked in the text to refer to the respective source. Depending on the field of study or type of work, it varies and appears as a numerical reference (IEEE style) or alphanumeric reference (AMS style, authorship trigraph). The following rules should be used in the final thesis (or just use the CGV template):

  • 1 author: the first three letters of the surname + year of publication... e.g. [Mei05]
  • 2 - 4 authors: the initial letters of the surnames (in the order they appear in the paper) + year of publication... e.g. [AB10], [XYZ15], [STUV12]
  • > 4 authors: the initials of the first 3 surnames, then a "+" sign, then the year of publication... e.g. [XYZ+04]
  • If there are multiple works by an author in the same year, lowercase letters are appended to the year... e.g. [Mei05a], [Mei05b]
  • If there are multiple sources for a text passage, they are separated by commas within a square bracket... e.g. [Mei05, XYZ+04]
  • If referring to a specific part of the source, this can be indicated in the reference list or citation bracket by specifying the page number... e.g. [Mei05, p.99].

4.2 Bibliography

In the numerical variant, sources in the bibliography are sorted according to their first appearance in the text. In the alphanumeric variant, they are sorted according to the contents of the bracket. The structure of an entry in the bibliography differs slightly depending on whether it is a conference paper or a book (chapter) (due to the different information to be provided). For example, a conference paper is structured as follows (according to the CGV scheme):

[XYZ99] LastnameInCapitalLetters1,␣Firstname1␣;␣Lastname2,␣Firstname2␣;␣Lastname3,␣Firstname3: Title of the Paper.␣In:␣Proceedings of the italicized conference on something Vol. X(Y),␣Location,␣Year,␣pp. <PageX-PageY>

When citing sources from the web, always include the URL/link and the date of retrieval. If possible, archive a copy of the internet source. Surveys/interviews are also sources. In this case the following information should be recorded:

[Mei15]␣Lastname1,␣Firstname1 (Interviewee)␣;␣Lastname2,␣Firstname2␣(Interviewer):␣Title of the Interview.␣Telephone/Personal/Written Interview/Conversation/Survey.␣Location,␣Date,␣Time

4.3 Citation Techniques

Exact (direct) quotation Useful for definitions and statements that could not be described more accurately. Placed in quotation marks if it is not longer than 4 lines. Otherwise, the entire quote block is indented (without quotation marks). The source is placed immediately after the quote in the text (Harvard method). Exact quotes must be honestly and accurately reproduced, without any rewording or distortions of the meaning. Text highlights or errors in the original text must also be reproduced (these can be marked with [sic] - Latin for "thus" or "really so"). Double quotation marks in the quote are replaced with single ones. Omissions are indicated by [...] (make sure that this does not distort the meaning of the original). Adaptations to the original, e.g. grammatical phrasing, should be written directly in square brackets at the relevant point - and also if words are added or highlights are made (write a clear comment in the square brackets, e.g. [emphasis added by the author]). Exact quotes should be used sparingly!

Paraphrased (indirect) quotation Here, the content of sentences or paragraphs from the original literature is reproduced in the same meaning as in the original text. The strict rules of exact quotation do not apply, but thoughts may not be altered, omitted, or added. In the sentence/paragraph that encompasses the content of the external source, there must be a "according to," "as per," etc. The citation bracket is placed before the period if the paraphrased quote only goes over one sentence. If the paraphrased quote is several sentences long, the citation bracket is placed after the period of the last sentence of the quote.

Figure citation Figures from external sources must be reproduced unchanged or the changes must be clearly indicated. The source reference belongs at the end of the figure caption (in square brackets). If a foreign illustration was used as a template for your own illustration, it also requires an indication, e.g. "according to [XY01, Fig. X.Y2]."

"Second-hand" quotes are those in which a source is cited, whose content represents a quote from the actual subject matter. Such quotes should be avoided! An exception is the unavailability of the original source, which is a rare case. A "second-hand" quote must be marked with the note "cited in" e.g. "[MXY+01] cited in [XY01]" (in this case, [MXY+01] would be the original source and [XY01] the cited source). Both works must be listed in the bibliography.

Good style is to mention the authors of an external source by name - if there are more than two authors, use the form "SurnameOfFirstAuthor et al." - however, it is also possible to use the citation bracket directly for this purpose. Examples:

Direct quote:

okay: In the study by [MYZ+01], it is described that these are " [...] crucial factors."

better: Meier et al. describe in their study that these are " [...] crucial factors." [MXY+01]

Indirect quote:

okay: According to [MS01], there are various crucial factors.

better: According to Meier and Schmidt, there are various crucial factors [MS01].

not so good: There are various crucial factors for this (see [MS01]).

back to table of contents

5 Literature Research

Valuable sources should be used, such as papers from well-known conferences with review systems or those that have been frequently cited. Printed sources are generally more credible and should be preferred over web sources such as forums, tutorials, or Wikipedia. Wikipedia can be a first point of reference, but it is scientifically controversial - it's better to search for "proper" literature from there. General problems include:

  • usually only one paper is given by the advisor
  • lack of overview over the field
  • there is a lot of literature and much of it is poor
  • there is not enough time for exhaustive research.

The strategy is to:

  • read the given paper completely (sometimes multiple times) and write down important technical terms (buzzwords)
  • read related work again to develop a sense of the field
  • examine related works (often it's enough to read the abstract and results/discussion to get an impression)
  • divide literature into relevant current works, overview articles (STARs) as a source pool, and older works (good candidates for backward search).

These resources (STARs, old works, buzzwords, names of major conferences) are the basis for further research. Where should one look?

5.1 Search engines (selection)

The SLUB has a subscription to many online portals, so you can download listed papers or books for free. However, this can only be done from the TUD IP address range (Uninetz). If you want to access it from home, you can use OpenVPN.

5.2 Techniques

Finding new works through older ones search for the older publication. Then show the works that cite the older work - this option is usually called "Referenced by" / "Cited by". Then research the displayed (new) works.

Finding new works through buzzwords enter buzzwords in the search engine. Sort by publication date and number of citations. Scan the first hits (possibly new buzzwords will emerge). Possibly search for research groups that are known in the specific field and search their publication directories.

Finding new works through conferences after finding relevant research areas and keywords, you can search for major conferences in these fields. Look at the lists of publications and then research them in more detail.

Finding new works through well known authors search for publication lists of authors who are frequently mentioned in the relevant field or whose names frequently appear in the bibliography. Note the order of authorship in the header of a paper. The first-named author is the author (of the largest part) of the work. The far right typically indicates the head of the department/institute/chair as the supervisor of the work.

5.3 Reading a Source I (Papers)

Begin with the Abstract and Results/Discussion section to quickly access the essential information. Ask the following questions about the source being examined:

  • What are the core contributions? (They are usually at the end of the Introduction)
  • What relevance do the core contributions have for your own work?
  • What results from cited publications are relevant to serve as justification for core contributions in other papers? Compile a list of these cited works and follow up on them.
  • What terminologies were introduced for the relevant core contributions? Are these terms interesting for your task? Can you expand your search queries using the new terms?

Don't panic if you don't understand everything immediately: Scientific papers usually contain highly condensed information. Typically, they need to be read multiple times to be fully understood.

5.4 Reading a Source II (Longer Works)

SQ3R method: survey, question, read, repeat, review

  • Survey: Get an overview, study the table of contents: What was covered? How is the text structured? What foundations does the author rely on? What is important for you?
  • Question: Formulate questions about the text - what information do you expect from this text for your own work?
  • Read: Read the relevant chapters for your question.
  • Repeat: It's okay if you don't fully understand everything on your first read of a chapter. Simply repeat what you understood (preferably aloud). If you get stuck: reread and repeat what you understood (preferably aloud). If you get stuck again: reread and...
  • Review: Summarize the content briefly in your own words. Were the questions about the text answered? Are there new questions?

5.5 How to rate a source

How do you know if the found source is useful? Even without specialized knowledge, you should pay attention to the following characteristics:

  • the work is current
  • the "related work" section is extensive
  • the contribution of the work is clearly highlighted in the introduction
  • the work has been cited frequently
  • the work was published in a major conference (if "ACM" or "IEEE" appears in the conference title, this is a good starting point)

back to table of contents

6 Writing Style

Always use scientific language - never use everyday language!

Technical terms

Use technical terms, but not to obscure content. The reader must be able to understand them. If unsure, create a glossary. If there is a technical term for something, use it instead of a synonym.

Fill-in phrases

Avoid standard phrases such as e.g.  "as can be easily seen...". Don't use relativisations ("many", "often", "mostly"), exaggerations ("enormous", "incredible"), filler words ("indeed", "well"), reassurance words ("somewhat", "somehow", "probably"), argument replacement words ("of course", "naturally"), or personal opinions. Your own statements are not prohibited, but must be critically reflected upon and justified. Stay humble in your explanations and avoid arrogant formulations (bad example: "The foundation is trivially provided by the well-known theories of tensor arithmetic"). Avoid formulations with "one" or "I".

Comprehensibility

Don't write artificially complicated, but as if you were orally explaining a scientific fact to a professor. Write concise, clear sentences that exclude ambiguities and are content-wise informative. Terms must be defined clearly and used in a consistent manner. Also strive for a consistent level of language. Stay logical and never lose the thread. Stay focused on the problem. Write for the reader! Guidelines for comprehensibility:

  • Each sentence contains a statement
  • Each paragraph contains a thought
  • Each section contains a group of thoughts

Examples:

Bad: It is a well-known problem in computer graphics that this interface limits the possibilities, which is why some data, such as textures, are not held in conventional main memory but are transferred to the graphics card and stored there in graphics memory.

Better: It is a well-known problem in computer graphics that this interface limits the possibilities. Therefore, it is common to store data such as textures in graphics memory instead of conventional main memory. This way, they only have to be transferred to the graphics card once.

Sentence Structure

Use subordinate clauses sparingly - avoid nested or unbroken sentences! Pay attention to clear role distribution: main information in the main clause, subordinate information in the subordinate clause. Eliminate subordinate clauses without (relevant) content. Avoid chains of genitives. Use verbs instead of nouns or auxiliary verb constructions (use "depends on" instead of "there is a dependency" or "is dependent on"). Whenever there is a choice between a verb and something else, choose the verb! Do not use too many prepositional phrases, meaning not more than one preposition ("in, under, over, between, in front of, after, against...") per sentence. Write in a positive sense instead of a negative sense - do not use double negations and write what is and not what is not. Also, avoid using too many passive formulations, but write in an active style.

Abbreviations

Use abbreviations sparingly and always use them unambiguously. Explain them at their first occurrence and create a list of abbreviations. Commonly known abbreviations (according to Oxford English Dictionary, Chicago Manual of Style, etc.) do not need to be included in the list. Do not rely on the reader to remember all abbreviations immediately: if you use formulas, constants, or abbreviations again many pages after their first introduction, explain them again with a brief repetition. For example, write "Here, the value α, which is the rotation angle, is used again to..." even though you introduced the variable α three chapters ago.

Structure

Avoid bullet point lists and write continuous text instead. Avoid frequent use of forward or backward references (e.g., "As will be seen in chapter X..." or "As shown in chapter Y...").

Figures

Use figures, tables, and diagrams to make complex textual statements more understandable, but avoid figures of trivial things. All figures or tables must be self-explanatory (axis labels, legends, color meanings, units, etc.). The use of a figure in no way makes a textual description obsolete: a) the body text must also be understandable without the figure and b) the figure must not replace the running text. Do not place essential new information solely in the figure (Negative example: The text describes that the effects shown in figure XY occur for these or those reasons. However, the effects themselves are only mentioned in the caption). All figures, tables, and diagrams must be referenced in the text.

Numbers

The lower numbers ("zero" to "twelve") are normally spelled out in text unless there is a particular emphasis on the numerical size. Units are generally abbreviated without a period (kg, km, h, min...). A narrow non-breaking space is placed between the value and the unit symbol (do not break the line here). Avoid writing unrelated numbers together (Negative example: "256 64-bit registers...").

Formulas

A consistent style should be used, and a uniform naming convention for variables, function names, etc. should be established. Do not use overloaded symbolism, but still strive for accuracy. Avoid using formulas directly in running text as much as possible, as they disrupt the reading flow and can sometimes sabotage the text layout; instead, use displayed formula environments. For example:

The Pythagorean theorem is a fundamental theorem of (Euclidean) geometry and states that:

a2 + b2 = c2

The equation holds for any right triangle, where a and b are its catheti, and c is its hypotenuse.

Do not unnecessarily include complicated and lengthy derivations in the main text. Reduce them to the essential points (and provide additional details in the appendix if needed).

Source code

Should never be included in its entirety in the main text! If necessary, only selected portions should be included due to special circumstances. Instead, explain the developed procedures using structure and flow diagrams or with pseudocode.

back to table of contents

7 Best Practices

In the following, you will find tips for the chapters presented according to the IMRAD (Introduction, Methods, Results and Discussion) scheme.

Introduction

At the very beginning of the written work, the introduction should provide a concise motivation for the task and elegantly introduce the topic. Here, the problem is presented in the context of an application, and the content of the work is briefly previewed. What problem was solved, why is it relevant? What is the approach? What was thematically limited or excluded? It is important to highlight your own contribution in a few concise sentences. (The conclusion should refer back to these contributions at the end to give the work a narrative bracket.) The introduction ends with an overview of the work — here, the contents of the individual chapters are briefly described (avoid trivial statements such as "In the results chapter 7, the results will be presented"). Hint: Avoid standard intros like "XY is an important field of application in computer graphics" or "XY is indispensable in computer graphics."

Related work

It should be shown who has already dealt with the topic or similar related topics, what solution approaches have been described and what the connection of the respective work to one's own is. Keep the "4 Questions" in mind as a mnemonic: What problem was tackled? How was the problem solved? What did it bring? How does it relate to your own work? In this chapter, it is particularly difficult to create a red thread and prevent the text from becoming a list of papers. Strategies that can be combined are available: chronological or aspect-oriented. In the chronological listing, related works are described in chronological order, giving a historical overview of the solution approaches to the problem. Typically, the first source in time is described in more detail, as well as the sources that follow more closely in time. Finally, the current state should be examined in more detail. The second strategy, aspect-oriented citation, involves dividing the papers into aspects of one's own problem. For example, if the topic is volume rendering with global illumination, papers on volume rendering in general should be presented first, then sources on advanced methods, and finally papers on the integration of global illumination, possibly even in separate sections.

Basics

Mathematical, technical, algorithmic, and other knowledge should be explained here, but only as much as is needed to understand the work. The author's own level of knowledge before starting the work can be considered as prior knowledge. More advanced basic knowledge or detailed mathematical derivations can be moved to the appendix. The work is not a textbook. Explanations of the technical terms used may be given here (or at the beginning of the methodology chapter).

Methodology and Implementation

Here, the author's own work is described conceptually, including problem analysis and solution search/finding. At this point, a theoretical examination of the material should take place — do not explain using concrete APIs or source code (pseudocode, however, is allowed). Only after that comes the description of the implementation — in most cases, the implementation as software. Pay attention to a clear and problem-oriented selection of code examples or (partial) class diagrams. Detailed and comprehensive presentations should be moved to the appendix if necessary.

Results

Here, an objective presentation of the results is made. A division into quantitative evaluation (generation of measurement data) and qualitative evaluation (surveys/expert feedback/description of peculiarities) is useful. For each evaluated issue, the result (e.g., as a figure or table) should be shown first, then described, and only then interpreted. Evaluation should not yet take place.

Discussion

Only in the discussion section are the results to be critically questioned and assessed. Based on this, the pros and cons (of the developed method) are worked out.

Summary

The work is briefly summarized and evaluated, the results/solutions are condensed in a conclusion (making a connection to the problem statements raised in the introduction). An assessment of the general usefulness of the developed methods can be given. Hint: End on a positive note! In the conclusion, show the "good" things first, then the "bad" ones. For the latter, note that they are solvable and that the developed methods are nevertheless promising.

Outlook

The outlook presents sensible extensions of the developed methods or possibilities for further research. Here, current weaknesses should be explained as opportunities for new concepts.

General

At the beginning of a main chapter, an overview of the following subchapters can be given. At the end of a main chapter, its content can be summarized and linked to the next main chapter. Anything that could hinder the reading flow (such as extensive tables, figures, mathematical derivations, or source code) should be moved to the appendix. Only really important snippets of source code should be included in the main part. It is better to avoid it and explain the underlying concepts/algorithms.

back to table of contents

8 Time Management

Week 1 2 3 4 5 6 7 8 9 10 11
Rough chapter structure, outline X                        
Research (read, structure, take note of relevant information) X X X X                  
Prototype implementation     X X X X X X X X      
Rough draft         X X X X          

Revision, correction, final draft

                 X X X    
Evaluation, creating graphics/tables, etc.                   X X    
Proof-reading                       X  
Printing, binding, submitting                         X

Figure 1 Schedule for a Bachelor's Thesis..

Create a schedule and stick to it (for example, as shown in Figure 1). Also plan buffer time for unforeseeable events and don't let holidays, such as Christmas and New Year's, or exam periods catch you off guard. The work should be evenly distributed throughout the week. A day off from work is important! Make daily to-do lists and plan enough breaks (more than 5 hours of concentrated work per day is hardly possible). Especially plan the written work in small manageable steps and in the evening try to complete one item from the list for the following day.

Start work at a set time every day, whether you feel like it or not! Don't forget the breaks (recommended are 15 minutes of break after 45 minutes of work).

Consider your biorhythm and don't schedule important tasks during times when you are in a "tired" phase (instead, do simple tasks: organize literature, take care of the household or relax...). Remember to engage in daily physical activity — it also promotes mental performance.

back to table of contents

9 Hints on the practical part/implementation

Especially in Bachelor's theses, you have a very limited amount of time for implementation. Therefore, focus on the essential core of your implementation. You don't need to reinvent the wheel, instead use existing software components and frameworks (e.g. for event handling, I/O operations, etc.). Your advisor can surely give you good hints, so that you don't have to develop your software from scratch. Discuss explicitly with your advisor the scope of functionality that your developed components should provide, and tackle these tasks first. If you realize during the course of your work that you still have time for further functionality, this will be welcome and credited to you as a bonus.

When implementing, always keep the KISS principle ( "Keep it simple and stupid" or "Keep it small and simple") in mind. Try to find the simplest implementation for the given problem. In general, a mature software product is not expected, a functional prototype is completely sufficient. However, don't skimp on comments in the source code and adequate documentation. Good program structure and software design are also essential. This helps everyone who wants to reuse your software later — for example, for follow-up theses, in teaching, etc. Perhaps you yourself may want to reuse your own software later and will be pleased to find that all functionalities are well-structured and explained.

back to table of contents

10 Procrastination Techniques

Many students invest too much time in the implementation. Reasons for this are sometimes "perfectionist thoughts" or the desire to include even more program features in the software. Unfortunately, there is also the problem that some prefer to engage in intensive implementation work to avoid the written work — possibly because they are not sure how to "get started". Most likely, they simply lack experience in planning and implementing such a large project.

The phenomenon of procrastination is not unknown and is also referred to as "student syndrome" among other things. If you yourself are plagued by such thoughts or feel unable or only idle to plan your activities, contact your supervisor in good time and develop a concept for work distribution together. Be aware: you are certainly not the first person to experience writer's block, and help is available!

back to table of contents

11 Presentation Guidelines

11.1 The Two Types

Intermediate presentation In this presentation, students should particularly show their current progress and receive feedback from the audience. Suggestions may arise that can improve the implementation of the task. The presentation should include an introduction to the topic and the presentation of related work to give the audience an overview of the topic. Then, the chosen approach should be justified, implementations shown, and current results presented. A slide showing a list of completed and uncompleted parts of the task should not be missing. For the unfinished tasks, a realistic estimation should be made with the help of a timetable of whether and how they can be completed on time.

Defense It is not just a summary of the contents produced. Due to the limited speaking time, students should demonstrate that they are able to select important content and leave out unimportant content (or only mention it on the sidelines). Above all, their own contributions should be emphasized. Try to present your knowledge in a way that is as understandable as possible. Unlike you, the listeners do not know every detail of the work and do not want to be informed about every little problem that occurred during the processing time. Start with a good motivation and introduce the task. Be creative and catch the interest of your listeners! This also includes explicitly listing the challenges (i.e., why is your problem a problem) once again. This can be done through a single slide entitled "Challenges." This is followed by a brief outline to provide a roadmap for the presentation. Then, related work and technical foundations should be presented first. These should be kept as brief as possible. A defense is not a lecture! In addition, many things have already been discussed in the intermediate presentation. The core of the work, their own achievements, should occupy most of the presentation time. For particularly extensive work, it may sometimes be necessary to discuss only one part in detail and present the remaining parts in an overview. Always keep the thread and consider carefully what knowledge the listener needs when to follow the presentation, thus avoiding duplicate explanations. Depending on the nature of the work, it can be helpful to start with an overall view (big picture) before explaining finer structures. This gives the listener a guide. Usually, results from measurements or surveys are presented and discussed at the end. The presentation ends with a brief conclusion in which they can emphasize their own achievement again and refer to the task statement/introduction. Here, you can tie things together and show the "Challenges" slide once again. This time, in addition to each sub-problem, briefly summarize your presented solution(s). This is followed by a demonstration of the developed application and the obligatory question and answer session.

11.2 Question and Answer session

Imagine the question and answer session as an opportunity to present more in-depth knowledge. Some questions may be directed towards you as an "expert" and may be quite detailed, while others may be simpler in nature and serve as an indication that parts of your presentation were not understood. React calmly and professionally in the latter case. Under no circumstances should you respond rudely or arrogantly — difficulties in understanding often come from the structure of the presentation (especially if many details are discussed but an overview is lacking). Do not view expert questions as a personal attack and try to answer as objectively as possible. This is not always easy, as you have spent a long time working on the task and may have become very attached to it. Try to approach the situation from a distance and do not justify your actions, especially if your approach is being questioned. Instead, allow for alternative solutions and provide a comparative assessment of your own approach.

11.3 Miscellaneous

Time Limit The given time limit is strict and must be strictly adhered to. During your presentation, the last 5 minutes of presentation time will be discreetly displayed to you. If you are about to exceed the time limit, you will be informed verbally. At this point, you should definitely wrap up the presentation, otherwise deductions in the evaluation may occur.

Clothes Dress appropriately for the presentations. However, you don't need to wear a suit. For defending a thesis or dissertation, a dark pair of long pants and a shirt or a plain sweater is sufficient. T-shirts with prints or casual pants can make you appear less credible as a presenter.

Performance Be as relaxed and confident as possible. Try not to appear too stiff (but also not too hyper). Use speech pace and accentuation to guide the presentation and direct the listener's attention.

Practice Practice your presentation beforehand — for example, in front of a mirror or with good friends. This way, you can already get some initial feedback on its comprehensibility. You will also learn to assess whether the presentation fits within the time limit. Practicing the presentation is also a great way to train a relaxed attitude and speaking style.

Slides Don't write everything you say on the slides. There is a great risk that it will appear as if you are simply reading off the slides. Instead, create good illustrations and diagrams. Slides support the presentation in this way much more effectively, as concepts are often more understandable and tangible when presented graphically. Avoid unnecessary text in your presentation.

Backup Slides If you had to remove detailed information from the presentation slides due to time constraints, don't hesitate to collect them as an appendix. In the Q&A session, the extra slides can help with your explanations and also show that you have a deep understanding of the subject matter.

References in Slides Adopted illustrations and descriptions of related work must be correctly marked as citations, e.g. with square brackets. Use the same abbreviations as in your written work. Create a slide with the references used in the presentation for the appendix, but do not show this slide during the presentation.

Intermediate Questions Generally, be open to spontaneous questions, but don't waste valuable speaking time. Try to return to the presentation as quickly as possible and refer to the Q&A session for questions that would take too long to answer.

back to table of contents

12 Thesis Evaluation Criteria

The Chair of Computer Graphics and Visualization pays attention to the following points in their assessments for final theses:

  • How extensive, how difficult, and how innovative is the work?
  • How much prior knowledge has the student brought in through lectures, exercises, or seminars?
  • Were the objectives of the task achieved, were objectives changed, were objectives expanded?
  • How is the student's working method with regard to goal orientation, prudence, systematicity, independence, and ability to discuss?
  • How is the written work regarding systematic structure, literature review, classification of the student's own work, comprehensibility, clarity, completeness, text and presentation quality?
  • How is the practical implementation regarding scope/effort, software design, completeness, stability, correctness, documentation, use of libraries and frameworks?

back to table of contents

13 Recommended Literature (German only)

  • Kornmeier, Martin: Wissenschaftlich schreiben leicht gemacht für Bachelor, Master und Dissertation. UTB (Haupt), 2013
  • Theuerkauf, Judith: Schreiben im Ingenieurstudium: Effektiv und effizient zu Bachelor-, Master- und Doktorarbeit. UTB Verlag, 2012
  • Ebel, Hans Friedrich: Bachelor-, Master- und Doktorarbeit: Anleitungen für den naturwissenschaftlich-technischen Nachwuchs. Wiley-VCH Verlag GmbH & Co. KGaA, 2009
  • Hohmann, Sandra: Wissenschaftliches Arbeiten für Naturwissenschaftler, Ingenieure und Mathematiker. Springer Vieweg, 2014
  • Balzert, Helmut ; Schäfer, Christian ; Schröder, Marion ; Kern, Uwe: Wissenschaftliches Arbeiten - Wissenschaft, Quellen, Artefakte, Organisation, Präsentation. W3l Verlag, 2008
  • Rechenberg, Peter: Technisches Schreiben: (nicht nur) für Informatiker. Carl Hanser Verlag GmbH & Co. KG, 2003
  • Prevezanos, Christoph: Technisches Schreiben: Für Informatiker, Akademiker, Techniker und den Berufsalltag. Carl Hanser Verlag GmbH & Co. KG, 2013
  • Weissgerber, Monika: Schreiben in technischen Berufen: Der Ratgeber für Ingenieure und Techniker: Berichte, Dokumentationen, Präsentationen, Fachartikel, Schulungsunterlagen. Publicis Publishing, 2010

back to table of contents

Fußnoten

  1. Note: The examples for formatting formulas may be difficult to read on your browser due to the TU Dresden corporate design. We apologise for this and recommend the PDF version (German only).

    ^

About this page

Ludwig Schmutzler
Last modified: Apr 17, 2023