Tips on writing a thesis

What goes into each part of an MSc thesis, and how to write, revise and polish it. By Stefan Sommer, with input from the Population Ecology Research Group.

Author

Stefan Sommer

Last revised on 11 June 2026.¹

A. What to write

Master’s theses in the natural sciences usually follow the IMRaD² structure: Introduction, Methods, Results, and Discussion. The introduction is preceded by a title, an abstract, and maybe keywords. The discussion is followed by acknowledgements, references, and probably an appendix.

1. Title

The title serves one purpose—to advertise.

Guidelines³

Ecology—Titles should be concise, informative, tell what the paper is about and what it found. It should contain key words necessary for digital search and retrieval methods. Avoid vague declarations (e.g., ‘effects of …’); strive for information content (e.g., ‘fungi kill tardigrades’).

Journal of Animal Ecology—[Use a] concise and informative title. Do not include the authorities for taxonomic names.

Tips

Create a title that is clear, concise, and captivating.
Be as general as possible to attract a large readership.
Be as specific as necessary to attract the expert readership.
Avoid abbreviations, acronyms, and metaphors.
Think twice about using humour; misplaced puns may compromise your scientific credibility.

2. Abstract

The abstract serves two purposes—to advertise and to summarise. The most common format is the single-paragraph abstract; it follows the IMRaD structure.

Guidelines

Ecology—The abstract should explain to the general reader why the research was done and why the results should be viewed as important. It should provide a brief summary of the research, including the purpose, methods, results, and major conclusions. Do not include literature citations in the Abstract. Avoid long lists of common methods or discursive explanations of what you set out to accomplish.

Journal of Animal Ecology—The Abstract […] should list the main results and conclusions, using simple, factual, numbered statements. Abstracts should normally include five points: 1. the background, 2. the goal of the study, 3. what was done in the study, 4. what was found and 5. what this means in a broad context. The Abstract should be written to be understood in isolation from the rest of the paper.

Tips

Write for a general readership.
Provide background information leading to the research question (or knowledge gap).
State the research question (or knowledge gap).
Explain how you tackled the research question (or filled the knowledge gap); describe only the concepts; omit the details.
State the main results.
Explain what the main results mean in a broader context.
Avoid citations, abbreviations, and acronyms (unless the acronyms are common knowledge; e.g., DNA, PCR).

3. Keywords

Keywords in a Master’s thesis serve no obvious purpose⁴.

Guidelines

Ecology—Following the Abstract, list between 6–12 key words/phrases. Each key word should be useful as an entry point for a literature search.

Journal of Animal Ecology—A list in alphabetical order not exceeding eight words or short phrases. More advice on selecting good keywords can be found here.

Tips

Use terms that appear in the title and abstract.
Restrict the length of individual terms to three words (‘snow vole’ and ‘life-history trait’ instead of ‘snow vole life-history trait’).
Avoid overthinking the keywords (or omit them).

4. Introduction

The introduction serves three purposes—to advertise, to summarise, and to contextualise.

Guidelines

Ecology—A brief Introduction describing the paper’s significance should be intelligible to the general reader of the journal. The Introduction should state the reason for doing the research, the nature of the questions or hypotheses under consideration, and essential background. The Introduction is not a place for a lengthy review of the topic.

Journal of Animal Ecology—This should state the reason for doing the work, the nature of the hypothesis or hypotheses under consideration, and should outline the essential background.

Tips

Structure the introduction as an inverted pyramid: Progress from the general to the specific.
Introduce the research area; if your research covers more than one area, introduce the most general one first and the most specific one last.
Identify a knowledge gap within your research area.
Explain why filling this knowledge gap is important.
State the aims (what you want to achieve).
State the objectives (how you want to achieve the aims); describe only the concepts; save the details for the methods section.
State hypotheses, expectations, or predictions (if applicable).

5. Methods

The methods section serves three purposes—to introduce the study system, to describe the data collection, and to explain the data analysis.

Guidelines

Ecology—The Methods section should provide sufficient information to allow someone to repeat your work. A clear description of your experimental design, sampling procedures, and statistical procedures is especially important. Do not describe commonplace statistical tests in Methods, but allude to them briefly in Results. If you list a product (e.g., animal food, analytical device), supply the name and location of the manufacturer. Give the model number for equipment specified.

Journal of Animal Ecology—Include sufficient details for the work to be repeated. Standard reporting guidelines should be followed where appropriate (examples include ROSES reporting standards for evidence synthesis and a checklist compiled by Ecology Letters for reporting standards in experimental studies). All data, program code, methods and research materials should be appropriately cited, see Reference section below for examples. Where specific equipment and materials are named, the manufacturer’s details (name, city and country) should be given so that readers can trace specifications by contacting the manufacturer. Where commercially available software has been used, details of the supplier should be given in brackets or the reference given in full in the reference list. Do not describe or refer to commonplace statistical tests in this section but allude to them briefly in Results.

Tips

Start with the description of the study system (field site, study organisms).
Continue with the description of the data collection (in chronological or otherwise logical order).
End with the description of the data analysis.
Use the past tense for describing the data collection and analysis.
Prefer the active voice if you did it (‘I sampled the population daily for two months.’).
Use the passive voice if others did it (‘The population was sampled every August from 2002 to 2018.’).
Use intuitive descriptors (‘the food and the temperature treatment’ instead of ‘treatments A and B’).
Describe all details that potentially affect the results.
Omit irrelevant details (usually, it doesn’t matter whether you analysed the data on a Mac or a PC; if it does matter, mention it).

6. Results

The results section serves one purpose—to provide new information.

Guidelines

Ecology—Results generally should be stated concisely and without interpretation, though in complex studies modest interpretation of individual parts can provide context helpful for understanding subsequent parts.

Journal of Animal Ecology—State the results and draw attention in the text to important details shown in tables and figures.

Tips

Order the results logically (e.g., from most important to least important or from basic to derived).
Present the results objectively.
Save interpretations for the discussion.
Use the past tense.
Be specific (‘survival increased’ instead of ‘survival changed’).
Use figures to illustrate trends and relationships.
Use tables to communicate precise values.
Arrange and number the figures and tables in the order in which you refer to them in the text.
Refer to all figures and tables in the text.

7. Discussion

The discussion serves one purpose—to convert the information into knowledge.

Guidelines

Ecology—The Discussion should explain the significance of the results. Distinguish factual results from speculation and interpretation. Avoid excessive review.

Journal of Animal Ecology—This should point out the significance of the results in relation to the reasons for doing the work, and place them in the context of other work.

Tips

Structure the discussion as an upright pyramid: Progress from the specific to the general.
Answer every research question raised in the introduction.
Explain what the results mean for your study system.
Discuss limitations; avoid being overly self-critical (do not put the limitations at the end of the discussion; end on a positive note).
Place the results into a wider context: How do the results of your study relate to the results of other studies? Do your results raise new questions?
End with the broadest conclusions that you can draw from your results (either as the last paragraph of the discussion or as a separate subsection headed Conclusions).

8. Acknowledgements

The acknowledgements serve one purpose—to credit where credit is due.

Guidelines

Ecology—Acknowledgments, including funding information, should appear in a brief statement at the end of the body of the text.

Journal of Animal Ecology—A brief statement acknowledging collaborators and research assistants who do not meet the criteria for authorship described above, or acknowledging funding sources, providing relevant permit numbers (including institutional animal use permits), or giving recognition to nature reserves or other organizations that made the work possible.

Tips

Thank everybody who contributed to your study; specify their contributions.
Acknowledge agencies that funded your work or issued permits.
Be creative (in a Master’s thesis, acknowledgements are less formal than in a published article).

9. References

The reference list serves one purpose—to direct readers to cited work.

Guidelines

Ecology—Avoid excessive citations; cite only essential sources. Before submitting the manuscript, check each citation in the text against the Literature Cited to see that they match exactly. […] The Literature Cited section of a paper may refer only to permanently archived material. If a reasonably diligent scholar 20 years in the future could not be assured of finding a particular source, it would not be acceptable as literature cited. Because Internet sources typically have a short half-life, they may not be included in Literature Cited sections unless there is reasonable evidence of permanency (e.g., Dryad). […] Do not list abstracts or unpublished material in the Literature Cited. These materials may be listed in the text as personal observations (by an author of the present paper), personal communications (information from others), public communications (information in published abstracts or information publicly distributed over the Internet but not permanently archived), unpublished manuscript, or unpublished data.

Journal of Animal Ecology—In text citations should follow the author–date method whereby the author’s last name and the year of publication for the source should appear in the text, for example, (Jones, 1998). The complete reference list should appear alphabetically by name at the end of the paper.

Tips

Use reference management software.
Pick your favourite citation style and stick to it.⁵

10. Appendix

The appendix serves one purpose—to provide material that disrupts the story of your thesis but might be of interest to the reader (e.g., preliminary analyses, raw data, R code).

Tips

Draft the appendix as carefully as the main document.
Design the figures and tables in the appendix like the figures and tables in the main document (i.e., use the same colours for the same types of results etc.).
Label appendix figures as Fig. A1, Fig. A2, etc., and appendix tables as Table A1, Table A2, etc.

B. How to write

Easy reading is damn hard writing.—Nathaniel Hawthorne

11. Outlining

First, write down the section headings 1 to 10 of this document. Next, write down the subsection and sub-subsection headings of your thesis; limit the heading depth to three levels (e.g., 3. Results, 3.1. Vital rates, 3.1.1. Survival). Finally, write down the first sentence (or just the topic) of each paragraph you plan to write. Avoid overthinking the outline. Once you start writing, the outline will evolve.

Alternatively, you may write down the answers to the following questions⁶:

What is the central question?
Why is this question important?
What data (variables) are needed to answer this question?
What methods are used to get those data?
What analysis must be applied for the data to answer the central question?
What data (values) were obtained?
What were the results of these analyses?
How did the analyses answer the central question?
What does this answer tell us about the broader field?

12. Drafting

Embrace the wisdom of professional writers:

The scariest moment is always just before you start.—Stephen King
Get it down. Take chances. It may be bad, but it’s the only way you can do anything really good.—William Faulkner
The first draft is just you telling yourself the story.—Terry Pratchett
Rewriting is the essence of writing well—where the game is won or lost.—William Zinsser

13. Revising

In the 1980s, writing experts at the consultancy Clearlines formulated a set of writing principles. A cofounder of Clearlines, George Gopen, later refined and incorporated these principles into his Reader Expectation Approach:

In trying to make of any English sentence the sense the writer intended, readers need to be able to answer four essential questions correctly. If the reader gets even one wrong, the interpretation arrived at will not be the one the writer intended. The four questions:

What is going on here?
Whose story is this?
How does this sentence connect backwards and forwards to its neighbors?
What word or words in this sentence should be read with special emphasis because they are the stars of this show?

Remarkably, the interpretive clues to the answers to all four of these essential questions are sent to the reader not through word choice, but rather through structural location: Where a particular word shows up in a sentence will control the way that word will be used by the reader. Simply stated, readers know where to look for what.⁷

George Gopen explained the reader expectation approach first in a paper and later, in more detail, in a book⁸. Before I describe this approach, I want to clarify some terms.

The sentence you just read consists of a subordinate clause (‘Before I describe this approach’) and a main clause (‘I want to clarify some terms’).
A main clause is also called an independent clause (because it could stand alone).
A subordinate clause is also called a dependent clause (because it cannot stand alone); Gopen calls it a qualifying clause (because it qualifies the main clause).
A clause consists of at least a subject⁹ and a verb.
A sequence of words that lacks subject and/or verb is called a phrase.¹⁰
The beginning of a sentence builds the topic position (topic positions are underlined). The topic position links backward and usually extends to the grammatical subject but not the verb. If the backward link appears after the verb, the topic position extends past the grammatical subject. In a multiclause sentence (e.g., the one you just read), each clause has its own topic position.
The end of a sentence builds the stress position (stress positions are boldfaced). The stress position begins when the reader feels that what is now being read is all that remains to be read. Naturally, this ‘feeling’ is somewhat subjective.¹¹ A stress position can be created by an ensuing period, colon, or semicolon; it cannot be created by an ensuing comma.
The ground a paragraph covers is the paragraph’s issue.
The claim a paragraph makes is the paragraph’s point.

Readers have rather fixed (and often unconscious) expectations¹² of where in a sentence, paragraph, or section certain kinds of information will appear. Readers expect¹³

that the action of a sentence will be articulated by its verb;
that a clause will tell the story of whoever or whatever shows up first as the grammatical subject;
that the grammatical subject will be followed almost immediately by the verb;
that the material most to be emphasised in the sentence will appear in the stress position;
that every stress position will be filled with material intended to receive significant emphasis;
that the material at the beginning of a sentence will provide a backward link to the previous sentence;
that the material at the beginning of a sentence will contextualise them for the new and important material to be presented later in the sentence;
that a paragraph will develop the issue stated at its beginning (within the first one to three sentences);
that the point of a section’s opening paragraph will arrive at the paragraph’s end;
that the point of most medial paragraphs will be given up front;
that the point of a section’s final paragraph will arrive at the paragraph’s end.

How to revise:

Express the action of a clause or sentence in its verb.
Make the doer of the action the subject of that verb.
Keep grammatical subject and verb together.
Put in the topic position at the beginning of a sentence that piece of old information that makes the appropriate link backward.
Put in the topic position the person or thing whose story a sentence is telling.
Put in the stress position that piece of new information that deserves the greatest emphasis.¹⁴
Fill every stress position with something stressworthy; ensure that everything stressworthy occupies a stress position.
State the issue of a paragraph at the paragraph’s beginning.
Violate any reader expectation when the violation has a good purpose.

14. Creating cohesion and coherence

Cohesion connects sentences; it creates a sense of flow.
Coherence connects ideas; it creates a sense of the whole.

Cohesion is established in the topic position by linking to the previous sentence. A common way of linking backward is to repeat a term and precede it with this, these, that, those, or such. Another way of linking backward is to state the relationship between sentences explicitly. If a sentence is ‘this’ in relation to the preceding sentence, start the sentence with ‘that’:

If a sentence is …	start it with …
an example	For example/For instance
another example	Moreover/Further/Additionally
a contrast or alternative	By contrast/Alternatively
a qualification or limitation	However/Yet
an inference or conclusion	Therefore/As a result/Consequently
a specification or clarification	Specifically/That is
a step in a (time) sequence	First/Second (Next)/Third (Finally)

Cohesion can be established by a technique termed topic changing¹⁵. In topic changing, the occupant of the stress position in one sentence becomes the occupant of the topic position in the next sentence:

	Topic position (old information)	Stress position (new information)
Sentence 1	A …	B
Sentence 2	B …	C	B is now old information
Sentence 3	C …	D	C is now old information

Topic changing can disrupt coherence, if the themes¹⁶ in the topic positions are unrelated. As a rule of thumb, limit a paragraph’s number of themes to three.

In a technique termed topic stringing¹⁷, the themes in the topic positions are related:

	Topic position (old information)	Stress position (new information)
Sentence 1	A …	B
Sentence 2	A …	C
Sentence 3	A …	D

Although topic stringing establishes coherence, it does not guarantee cohesion.

To create paragraphs that are both cohesive and coherent, change and string topics. To establish cohesion across paragraphs, either signal forward at a paragraph’s end or link backward at a paragraph’s beginning—or do both.¹⁸

15. Diagnosing paragraphs

In each sentence,

highlight the grammatical subject in blue;
highlight the verb in red;
underline the action;
italicise the agent;
[bracket the topic position];
boldface the stress position;
‘put in quotation marks the key piece of new information’.

[In clear, cohesive, and coherent paragraphs, the blue words] are ‘italicised’; the red words are ‘underlined’; and the bold words are ‘in quotation marks.’¹⁹ That is, the agents are represented by the subjects; the actions are expressed by the verbs; and the key pieces of new information are put in the stress positions. Also, the topic positions link backward; the verbs cling to the subjects; and the subjects span maximally three themes.

16. Tips on style

Style matters for at least three reasons: It gets your message across; it earns you trust; and it adds beauty to the world.²⁰

Voice and tense

Prefer the active voice.
Use the passive voice when²¹
- it is the only way to move the backward-linking old information to the topic position;
- it is the only way to move the person or thing whose story it is to the topic position;
- it is the only way to move the most important new information to the stress position;
- the story is passive and benefits from the verb indicating that passivity. (‘Pollen is dispersed by bees’ is a passive story about pollen. ‘Bees disperse pollen’ is an active story about bees.²²)
Control the tenses:

For each time (past/present/future), there are four tenses. The first is for actions occurring in the moment or for which duration is not important: I wrote, I write, I will write (simple past, present, and future). The second is for actions ongoing through time: I was writing, I am writing, I will be writing (past, present, and future progressive). The third is for completed actions: I had written, I have written, I will have written (past, present, and future perfect). Finally, the fourth combines the ‘passage of time’ and ‘completed’ implications: I had been writing, I have been writing, I will have been writing (past, present, and future perfect progressive).²³

Punctuation²⁴

Place a serial (aka Oxford) comma before the conjunctions and and or in a list of three or more items. (‘We distinguished four life stages: newborn, juvenile, subadult, and adult.’)
Distinguish between hyphen (-), en dash (–), and em dash (—):
- A hyphen connects certain compound words, particularly adjectives (long-term study, state-of-the-art methods).
- An en dash indicates a range (20–25 °C, 2010–2019²⁵) and connects coequal words in compound terms (source–sink dynamics, human–wildlife conflict).
- A single em dash functions like a colon—it separates two parts of a sentence.
- A pair of em dashes—like a pair of commas (or parentheses)—demarcates an insertion in the middle of a sentence.

Sentences, clauses, phrases, and words

Beware of excessively long sentences; they are difficult to phrase clearly.
Avoid long phrases and clauses before the main clause. (‘Because I was unable to identify the reproductive type of experimental rotifers that died before they carried either mictic or amictic eggs externally, I assumed… .’)
Use parallel phrasing²⁶ to describe similarities and contrasts (‘whereas the highland population increased, the lowland population decreased’ instead of ‘whereas the highland population increased, the population in the lowlands declined’).
Avoid long strings of prepositional phrases (‘I studied the dynamics of rotifer populations from two pollution periods of a subalpine lake in northern Italy under controlled conditions in the laboratory at the University of Zurich.’).
Use affirmative rather than negative phrasing (‘juveniles died’ instead of ‘juveniles did not survive’)—unless you want to emphasise the do not aspect.
Do not join independent clauses by a comma.
Do not mix UK and US spelling.
Avoid long abstract nouns. (‘I developed a strategy to manage and conserve biodiversity’ instead of ‘I developed a biodiversity conservation management strategy.’)
Avoid meaningless words (actually, really, very) and redundant phrases (each and every, every single, first and foremost, last but not least).
Minimise the use of abbreviations and acronyms.
Define abbreviations and acronyms at first use—unless they are common knowledge (DNA, PCR).
Spell out that is (i.e.) and for example (e.g.) outside parentheses.
Spell da wurds correctly.

C. What to read

Gopen GD (2004) The Sense of Structure: Writing from the Reader’s Perspective. Pearson, New York, NY.
Heard SB (2026) The Scientist’s Guide to Writing: How to Write More Easily and Effectively throughout Your Scientific Career. 3rd Edition. Princeton University Press, Princeton, NJ.
Pinker S (2014) The Sense of Style: The Thinking Person’s Guide to Writing in the 21st Century. Penguin, New York, NY.

Good prose—on the sentence level, the paragraph level, and most often on the document level as well—tends to demonstrate three consistent characteristics:

Nothing arrives that the reader cannot handle at the moment of its arrival.
Everything leans forward.
Everything actually goes into one of the directions it had been leaning.

George Gopen, The Sense of Structure

Footnotes

Drafted in 2019 with input from Arpat Ozgul and the then-members of the Population Ecology Research Group: Cecilia Bardier, Gaurav Baruah, Dominik Behr, Eva Conquet, Gabriele Cozzi, Dilşad Dağtekin, Valentin Graf, Genoveva Haibach, Natasha Harrison, David Hofmann, Nicholas James, Dechen Lham, Steffen Mink, Manja Schleich.↩︎
Heard (2026, p. 76).↩︎
Author guidelines of the journals Ecology and Journal of Animal Ecology were accessed on 04 November 2019. The guidelines have since been updated.↩︎
Keywords in publications have experienced a shift in purpose: Before the advent of full-text search engines, they were used mostly to index manuscripts; nowadays, they are used mostly to assign reviewers.↩︎
If you are writing a manuscript for publication, follow the author guidelines of your target journal.↩︎
Quoted from Heard (2026, p. 68).↩︎
This text was originally copied from Gopen’s website. The website has since been updated.↩︎
Gopen (2004); see section C. What to read.↩︎
In imperative sentences (e.g., the tips in this document), the subject is only implied: ‘Double check the references.’ (‘You, double check the references.’)↩︎
For details on sentences, clauses, phrases, and more, see Pinker (2014, ch. 4).↩︎
The stress position starts when you would raise the voice if you read the sentence out loud.↩︎
The reader expectations are quoted, and occasionally paraphrased, from Gopen (2004).↩︎
‘Readers expect’ is short for ‘most contemporary readers of English expect most of the time.’↩︎
For how to use artificial emphasis, see Gopen (2004, pp. 48–49).↩︎
Gopen (2004, pp. 72–75).↩︎
Here, a theme is a group of related terms (often the sentences’ subjects). For example, the terms survival rate, maturation rate, and reproduction rate might build the theme vital rates; and the terms females, juveniles, and dispersers might build the theme study individuals.↩︎
Gopen (2004, pp. 76–77).↩︎
For details, see Gopen (2004, pp. 134–143).↩︎
In this example, the ‘action’ (underlined) is passive.↩︎
Paraphrased from Pinker (2014, prologue).↩︎
Gopen (2004, p. 153).↩︎
The examples are from Gopen and Swan (1990) The science of scientific writing. American Scientist 78, 550–558.↩︎
Quoted from Heard (2026, p. 174).↩︎
For extensive discussions, see Gopen (2004, pp. 157–233) and Pinker (2014, ch. 6).↩︎
In text outside parentheses, spell out the en dash: ‘from 20 to 25 °C’, ‘from 2010 to 2019’.↩︎
Structural parallelism is explained in detail in Pinker (2014, ch. 4).↩︎