Scientific Writing: The Complete Guide to Theses, Research Papers, Reviews, Systematic Reviews and Meta-analysis
A step-by-step guide for students, researchers and doctors: find a research gap, search the literature, write each section, run a systematic review and meta-analysis, use AI tools ethically and publish in good journals, with a free 56-video course in Urdu/Hindi.
Subscribe on YouTubeGood scientific writing starts long before the first sentence: with a clear research gap, focused questions and a systematic literature search. Then you write in a fixed structure (IMRaD for papers, chapters for a thesis, PRISMA for systematic reviews), revise hard, choose the right journal and respond carefully to reviewers. My free Scientific Writing Zero-to-Hero playlist teaches every step in Urdu/Hindi; give it 2 to 5 hours a day and you can write your first paper with confidence.
| Free course | Scientific Writing Zero-to-Hero playlist: 56 videos, from research gaps to journal submission |
|---|---|
| Masterclass | The complete Scientific Writing Master Class recording (2 h 46 min) |
| Free course with certificate | Using AI tools in Scientific Writing (about 4 hours, all levels) |
| Paid course | Systematic Review and Meta-analysis: two months, self-paced, with R, Excel, Word and EndNote |
| Language | Urdu/Hindi, with English scientific terms |
| Who it is for | Bachelor’s, Master’s, MPhil and PhD students, postdocs, lecturers, professors and medical doctors |
Most students are never taught how to write science. They are asked to “write a review” or “start the thesis” and left alone with a blank page, hundreds of PDFs and a supervisor who wants results. The result is months lost, rejections, and sometimes plagiarism problems that could have been avoided.
In this guide I share the system I teach in my courses and use in my own research, from the first idea to the accepted paper. It covers the five documents researchers write most: a thesis, an original research paper, a narrative review, a systematic review and a meta-analysis.
In this guide
- What scientific writing is, and the 5 document types
- Research gap, questions and keywords
- Search and read the literature
- Write an original research paper (IMRaD)
- Write a thesis
- Write a narrative review
- Systematic reviews and meta-analysis
- Use AI tools ethically
- Choose a journal, submit and publish
- A learning plan with the free course
- All 56 videos
- Links and related guides
- Key terms
- FAQ
What is scientific writing, and which type of document should you write?
Scientific writing is clear, precise and evidence-based writing that reports research so others can understand, check and build on it. The five most common documents are a thesis (your full degree research), an original research paper (one new study), a narrative review (an expert summary of a field), a systematic review (a reproducible search and appraisal of all studies on one question) and a meta-analysis (a statistical combination of the results of those studies).
| Document | Purpose | Method | Reporting guide | Typical time |
|---|---|---|---|---|
| Thesis | Show you can do independent research for a degree | Chapters: introduction, literature review, methods, results, discussion | Your university’s guidelines | Months to years |
| Original research paper | Report one new study and its findings | IMRaD: Introduction, Methods, Results, Discussion | Study-specific guides on the EQUATOR Network | 2–6 months after data |
| Narrative review | Summarise and interpret a field, find gaps | Expert reading and structured themes | Journal’s author guidelines | 2–4 months |
| Systematic review | Answer one focused question from all available evidence | Protocol, systematic search, screening, risk of bias | PRISMA 2020 | 6–12 months |
| Meta-analysis | Combine results statistically into one estimate | Effect sizes, pooling, heterogeneity, publication bias | PRISMA 2020 (with the systematic review) | Added to the systematic review |
In the first lessons of the course I explain what scientific writing is, why it matters for your career, and a basic plan to start writing even when you feel stuck.
- Pick the document type first: thesis, paper, narrative review, systematic review or meta-analysis.
- Each type has its own structure and reporting guideline.
- Writing is the last step of a well-planned process.
How do you find a research gap and turn it into research questions?
Read 5 to 10 recent review articles in your field, note what they say is unknown, contradictory or understudied, and check whether newer papers have already filled that gap. Turn the gap into one to three specific research questions (for systematic reviews, use the PICO format), then extract keywords and synonyms from those questions to build your search.
- Step 1Read reviews5–10 recent reviews in your area.
- Step 2Spot the gapUnknowns, conflicts, missing methods or populations.
- Step 3Check it is still openSearch the newest papers on that gap.
- Step 4Write questionsOne to three specific, answerable questions.
- Step 5Build keywordsMain terms, synonyms and spelling variants.
| Type of research gap | What it means | Example question |
|---|---|---|
| Evidence gap | Studies disagree or results conflict | Why do studies report opposite effects of X on Y? |
| Knowledge gap | A topic has simply not been studied | What is the effect of X in population Z? |
| Methodological gap | Better methods could change the answer | Does a newer method confirm earlier results? |
| Population gap | A group, region or species is missing | Does the effect hold in South Asian populations? |
| Theoretical gap | No model explains the findings | Which mechanism explains the effect of X? |
| Practical gap | Research has not been applied in practice | Can the method work in low-resource settings? |
Use PICO for focused questions
For clinical, health and many biological questions, the PICO format makes a question precise and searchable: Population, Intervention (or exposure), Comparison and Outcome. The Cochrane Handbook explains how to define each part for a systematic review.
Example: from research gap to PICO question and keywords
Research gap: Reviews report mixed effects of plant growth-promoting bacteria (PGPB) on wheat under drought. PICO question: P (population): wheat plants grown under drought stress I (intervention): inoculation with plant growth-promoting bacteria C (comparison): non-inoculated plants O (outcome): grain yield, biomass and relative water content Research question: Does inoculation with PGPB increase grain yield of wheat under drought stress compared with no inoculation? Keywords and synonyms: PGPB: "plant growth-promoting bacteria", PGPB, PGPR, "plant growth-promoting rhizobacteria", bioinoculant* drought: drought, "water stress", "water deficit" wheat: wheat, "Triticum aestivum" yield: yield, biomass, "relative water content"
These lessons show the full process, including the six types of research gap, how to speed up gap-finding and how to turn a gap into research questions from microbiology and bioinformatics examples:
- Gaps come from reading recent reviews critically.
- Write specific, answerable questions; use PICO for systematic reviews.
- Keywords and synonyms come directly from the questions.
How do you search and read the scientific literature efficiently?
Search trusted databases such as Scopus, Web of Science, PubMed and Google Scholar with a Boolean query built from your keywords, save every result in a reference manager such as EndNote or Zotero, and read in passes: title and abstract first, then figures and conclusions, and the full text only for the most relevant papers. AI tools can help you discover and summarise papers, but always read and cite the original.
| Database | Best for | Access |
|---|---|---|
| Scopus | Broad coverage of peer-reviewed journals in all fields, citation analysis | Through your university |
| Web of Science | Curated high-quality journals, citation tracking | Through your university |
| PubMed | Medicine, health and life sciences, with MeSH terms | Free |
| Google Scholar | Quick broad searching, theses and grey literature | Free |
Combine synonyms with OR, concepts with AND, use quotation marks for phrases and * for word endings. The same idea works in every database; only the field codes change:
Scopus:
TITLE-ABS-KEY ( ( "plant growth-promoting bacteria" OR pgpb OR pgpr OR bioinoculant* )
AND ( drought OR "water stress" OR "water deficit" )
AND ( wheat OR "Triticum aestivum" ) )
Web of Science:
TS=( ("plant growth-promoting bacteria" OR PGPB OR PGPR OR bioinoculant*)
AND (drought OR "water stress" OR "water deficit")
AND (wheat OR "Triticum aestivum") )
PubMed:
("plant growth-promoting bacteria"[tiab] OR PGPR[tiab]) AND (drought[tiab] OR "water stress"[tiab]) AND (wheat[tiab] OR "Triticum aestivum"[tiab])
Manage references from day one
Never type references by hand. Export search results to a reference manager such as EndNote or the free Zotero, keep one library per project, and insert citations with the Word plugin so your reference list updates itself. My EndNote crash course covers this from zero.
Read smarter, not longer
Read in three passes: first the title, abstract and conclusion; then the figures, tables and headings; and only then the full text of papers you will really use. Keep a simple literature table (author, year, method, main finding, limitation) as you go. Discovery tools such as Connected Papers and ResearchRabbit show papers related to a key article.
- Search Scopus, Web of Science, PubMed and Google Scholar with Boolean queries.
- Use EndNote or Zotero from the first day.
- Read in passes and keep a literature table.
How do you write an original research paper?
Use the IMRaD structure: Introduction (why the study matters and the gap it fills), Methods (exactly what you did, so others can repeat it), Results (what you found, with figures and tables, without interpretation) and Discussion (what the results mean, how they compare with other studies, limitations and conclusions). Write the methods and results first, then the discussion and introduction, and the title and abstract last.
| Section | What it answers | Tips |
|---|---|---|
| Title | What is the paper about? | Specific and searchable; include the main variables and organism or population |
| Abstract | What did you do and find, in 200–300 words? | Background, aim, methods, key results with numbers, conclusion |
| Introduction | Why does this study matter? | Funnel from broad context to the specific gap, end with your aim or hypothesis |
| Methods | What exactly did you do? | Enough detail to repeat the study: design, materials, procedures, statistics |
| Results | What did you find? | Follow the order of the methods; let figures and tables carry the data |
| Discussion | What does it mean? | Main finding first, compare with literature, explain, limitations, future work |
| Conclusion | What should the reader remember? | Two to four sentences, no new data |
- 1stFigures and tables. Decide the story your data tell.
- 2ndMethods and results. The easiest sections, because they describe what you did and found.
- 3rdDiscussion. Interpret the results against the literature.
- 4thIntroduction. Now you know exactly which gap your results fill.
- LastTitle, abstract and keywords. Summarise the finished paper.
Before you write, check the reporting guideline for your study design on the EQUATOR Network (for example CONSORT for randomised trials or STROBE for observational studies). Reviewers and editors use these checklists.
Abstract template (fill in the brackets)
Background: [One or two sentences on the problem and why it matters.] However, [the research gap]. Aim: This study aimed to [main objective]. Methods: We [design] in [population/organism/setting, sample size]. [Main measurements] were analysed using [key statistical method]. Results: [Main finding with numbers, e.g. X increased Y by 25% (p < 0.01)]. [Second key finding]. Conclusion: [What the results mean and their main implication]. [One sentence on future work or application].
The two-part Zero-to-Hero course (over 16 hours together) walks through writing a paper section by section, abstracts, EndNote, journal selection and the full submission process.
- Use IMRaD and the right reporting checklist.
- Write methods and results first, the abstract last.
- Let figures and tables tell the story.
How do you write a thesis?
Find out first whether your university expects a monograph (one long document) or an article-based thesis (published papers with an introduction and a general discussion). Then make a chapter outline with your supervisor, write the literature review and methods early, turn each results chapter into a paper if you can, and finish with the general introduction, discussion and abstract.
| Chapter | Contents |
|---|---|
| Abstract | The whole thesis in one page |
| 1. Introduction | Background, problem, research gap, aims, objectives or hypotheses |
| 2. Literature review | What is known, what is not, and how your work fits; often a review paper in an article-based thesis |
| 3. Materials and methods | All methods in detail (or methods inside each paper chapter) |
| 4–6. Results chapters | One chapter per objective or per published paper |
| 7. General discussion | How the chapters answer the research questions together, limitations and future work |
| References and appendices | Managed in EndNote or Zotero, in your university’s style |
- Get the rules early: format, word limits, citation style and whether published papers can be included.
- Outline before writing: one page with chapter headings and two to three bullets each, agreed with your supervisor.
- Write every day: short, focused sessions beat rare marathons. Protect two hours of deep work.
- Turn chapters into papers: a literature review chapter can become a review paper, and each results chapter an original paper.
- Get feedback in small pieces: send one section at a time, not the whole thesis at the end.
Writing a thesis is easier with a good mentor and community; I explain why mentorship matters in the course. Planning a PhD abroad? Read my guides on studying abroad for a PhD and research and Marie Curie PhD positions.
- Check monograph vs article-based thesis rules first.
- Outline with your supervisor, then write daily.
- Turn chapters into publishable papers.
How do you write a narrative review paper?
Choose a focused topic with a clear gap, search the main databases, read and organise the papers into themes in a literature table, build an outline around those themes, and write a critical synthesis: not a list of studies, but what the evidence shows together, where it disagrees and what research should come next. Finish with a strong conclusion and future perspectives section.
Pick a focused topic with a gap
A review on “plant microbiomes” is too broad; “bacterial inoculants for drought tolerance in cereals” is a review you can finish. Check that no recent review already covers the same angle.
Search, screen and organise
Search Scopus and Web of Science, save results in EndNote, and record each relevant paper in a literature table. Group the papers into three to six themes.
Outline by themes
Each theme becomes a section with subheadings. Plan one summary table and one or two figures, such as a mechanism diagram or a graphical abstract.
Write a critical synthesis
Compare, contrast and explain: which findings agree, which conflict and why (methods, conditions, populations). End every section with what is still unknown.
Conclude with future perspectives
Readers and editors value a clear list of research priorities. This is often the most cited part of a review.
- Narrow the topic until the review is finishable.
- Organise papers into themes with a literature table.
- Synthesise critically and end with future perspectives.
How do you conduct a systematic review and meta-analysis?
Define a focused PICO question, write and register a protocol (for example on PROSPERO), search several databases with a documented search string, screen titles, abstracts and full texts against fixed criteria with two reviewers, extract data, assess risk of bias, and synthesise the results. If studies are similar enough, a meta-analysis pools their effect sizes with a fixed- or random-effects model, checks heterogeneity and publication bias, and the review is reported with the PRISMA 2020 checklist and flow diagram.
Ask a focused question
Use PICO (or a variant such as PECO for exposures). Define inclusion and exclusion criteria before you search.
Write and register a protocol
Describe your methods in advance and register health-related reviews on PROSPERO; other fields can register on the Open Science Framework. Registration reduces bias and duplicate work.
Search systematically
Search at least two or three databases (for example Scopus, Web of Science and PubMed) with the same logic, and record the date and full search string for each.
Screen in two stages
Remove duplicates, then screen titles and abstracts, then full texts, ideally with two independent reviewers. Tools such as Rayyan make this faster. Record numbers for the PRISMA flow diagram.
Extract data
Use a piloted extraction sheet: study details, population, intervention, outcomes, sample sizes, means, standard deviations or event counts.
Assess risk of bias
Use a validated tool for each design, such as RoB 2 for randomised trials, ROBINS-I for non-randomised studies of interventions or the Newcastle-Ottawa Scale for observational studies.
Synthesise and meta-analyse
Calculate effect sizes (for example risk ratio, odds ratio, mean difference or standardised mean difference), pool them, and show the results in a forest plot.
Rate certainty and report
Rate the certainty of evidence with GRADE (high, moderate, low or very low) and report with the PRISMA 2020 27-item checklist and flow diagram.
| Meta-analysis term | What it means |
|---|---|
| Effect size | A number that shows the size and direction of an effect in one study, such as a risk ratio or mean difference |
| Fixed-effect model | Assumes all studies estimate the same true effect |
| Random-effects model | Allows the true effect to vary between studies; usually more realistic |
| Heterogeneity (I²) | How much results vary beyond chance; the Cochrane Handbook gives rough guides from 0–40% (might not be important) to 75–100% (considerable) |
| Forest plot | Shows each study’s effect and confidence interval, and the pooled result |
| Funnel plot and Egger’s test | Check for small-study effects and possible publication bias |
| Subgroup analysis and meta-regression | Explore why results differ, for example by dose, region or study quality |
A meta-analysis can be run in free software. Here is a minimal example with the R package metafor, using its built-in BCG vaccine dataset:
install.packages("metafor") # once
library(metafor)
# log risk ratios and their variances for each study
dat <- escalc(measure = "RR", ai = tpos, bi = tneg, ci = cpos, di = cneg, data = dat.bcg)
res <- rma(yi, vi, data = dat) # random-effects model (REML)
summary(res) # pooled effect, I² and heterogeneity test
forest(res, atransf = exp) # forest plot on the risk-ratio scale
funnel(res) # funnel plot
regtest(res) # Egger's test for funnel plot asymmetry- Systematic searching, screening and PRISMA reporting
- Meta-analysis in R and MS Excel
- Writing in MS Word with EndNote
- From question to a paper you can submit to a good journal
- Register a protocol before you search.
- Screen with fixed criteria and report with PRISMA 2020.
- Meta-analysis pools effect sizes; always check heterogeneity and bias.
How can you use AI tools in scientific writing ethically?
Use AI to search, summarise, organise ideas, check grammar and improve clarity, but write the science yourself, verify every fact and reference against the original paper, and disclose how you used AI as your target journal requires. AI tools cannot be listed as authors, because they cannot take responsibility for the work. Never paste confidential or unpublished data into tools that may store it.
| Good uses of AI | Risky or unacceptable uses |
|---|---|
| Finding and mapping related papers | Citing papers you have not read, or AI-invented references |
| Summarising a paper you then read yourself | Copying AI text into your paper without checking or editing |
| Improving grammar, clarity and flow | Generating data, results or images that look real |
| Brainstorming outlines and keywords | Listing an AI tool as a co-author |
| Checking a draft against a reporting checklist | Uploading confidential manuscripts you are reviewing |
Publishing bodies agree on the core rules. The ICMJE recommendations say AI tools should not be listed as authors and that authors should disclose AI use; the Committee on Publication Ethics (COPE) takes the same position. Always read your target journal’s own AI policy.
See more tools in my guides to AI tools for students and academics and the prompt engineering guide.
- AI helps with search, structure and language, not with the science itself.
- Verify every reference; AI can invent citations.
- Disclose AI use and never list AI as an author.
How do you choose a journal, submit your paper and respond to reviewers?
Shortlist journals that publish similar papers, check their scope, indexing (Scopus or Web of Science), impact, fees and review time, and avoid predatory journals with the Think. Check. Submit. checklist. Follow the author guidelines exactly, write a short cover letter, and when reviews arrive, answer every comment politely and point by point, showing exactly what you changed.
- Look at your reference list: journals you cite most often are good targets
- Use journal finders such as Elsevier Journal Finder with your title and abstract
- Check indexing and ranking in Scopus, Web of Science or SCImago Journal Rank
- Avoid predatory journals with the Think. Check. Submit. checklist
- Get an ORCID iD so all your papers are linked to you
- Agree on authorship early using the ICMJE authorship criteria
Cover letter template
Dear Dr. [Editor's name], Please consider our manuscript entitled "[Title]" for publication as a [research article / review] in [Journal name]. [One or two sentences on the problem and the research gap.] In this study, we [what you did]. We found that [main finding]. These results [why they matter for the journal's readers]. This work fits the scope of [Journal name] because [reason]. The manuscript has not been published and is not under consideration elsewhere. All authors have approved the manuscript and agree with its submission. [Any AI-use or conflict-of-interest statement required by the journal.] Thank you for your consideration. Sincerely, [Name, affiliation, email, ORCID]
Response to reviewers template
Reviewer 1, Comment 1: [Copy the comment exactly.] Response: Thank you for this helpful comment. We agree and have [what you changed]. [If you disagree: We respectfully note that ... because ... (with a reference).] Change in manuscript: Page [x], lines [y–z]: "[new or revised text]". Reviewer 1, Comment 2: ...
After acceptance, share your work: a conference poster or a short presentation helps your paper reach more readers. See my guide to making conference posters with AI tools.
- Choose journals by scope, indexing and fit, never by speed alone.
- Avoid predatory journals with Think. Check. Submit.
- Answer every reviewer comment point by point.
What is the best way to learn scientific writing with the free course?
Watch the Scientific Writing Zero-to-Hero playlist in order and give it 2 to 5 hours a day. Start with the short introduction lessons, then work through research gaps, literature searching and the full courses, and do every assignment on your own topic. Add the free AI tools course with a certificate, and join the Systematic Review and Meta-analysis course when you are ready to publish one.
| Group | Videos | Time | What you learn |
|---|---|---|---|
| Start here | 8 | 1 h 41 min | What scientific writing is, a basic plan, its importance, free resources and the community |
| Full courses, masterclasses and workshops | 5 | 22 h 12 min | Two full Zero-to-Hero courses, the masterclass and two workshops |
| Research gap, questions and keywords | 14 | 1 h 48 min | Finding research gaps, writing research questions and building keywords |
| Searching and reading the literature | 9 | 2 h 16 min | Searching Scopus and Web of Science, reading papers and free literature tools |
| References and writing in MS Word | 3 | 3 h 17 min | EndNote, word clouds and writing proposals in MS Word |
| AI tools, plagiarism and AI detection | 8 | 41 min | AI writing tools, plagiarism, AI detection and responsible use |
| Time management and mentorship | 4 | 17 min | Planning your time, focused work and why a mentor matters |
| Posters, slides and sharing your research | 5 | 1 h 59 min | Conference posters and slides with Claude Code, and research opportunities |
- Week 1Foundations. Start here group, plus the masterclass recording.
- Week 2Gap and questions. Find your gap, write questions and keywords; do the assignments.
- Week 3Literature. Search Scopus and Web of Science, set up EndNote, build your literature table.
- Weeks 4–6Write. Zero-to-Hero Part 1 while drafting your review or paper section by section.
- Weeks 7–8Submit. Zero-to-Hero Part 2: journal selection, cover letter and submission.
- Give the playlist 2 to 5 hours a day.
- Work on your own topic while you learn.
- Add the free AI tools course and, when ready, the SR & MA course.
All 56 videos of the Scientific Writing course
Start here 8 videos · 1 h 41 min
- 1
- 2
- 3
- 23Importance of Scientific WritingVideo · 22:04
- 21
- 22
- 24
- 31Future outlook for Scientific Writing SeriesVideo · 8:00
Full courses, masterclasses and workshops 5 videos · 22 h 12 min
- 9
- 10
- 12Scientific Writing Master Class at CodanicsVideo · 2:46:27
- 20
- 44
Research gap, questions and keywords 14 videos · 1 h 48 min
- 26How to find a research gap?Video · 34:52
- 27Six types of research gap in scientific writingShort · 0:44
- 28Read 5 to 10 to find research gapVideo · 6:58
- 29
- 32
- 33
- 34
- 35
- 36
- 37
- 38Research Questions from a Bioinformatics PerspectiveVideo · 11:18
- 39Defining Keywords from Research QuestionsVideo · 15:17
- 7
- 8Student Assignment: Designing Keywords for a Review PaperAssignment · 0:30
Searching and reading the literature 9 videos · 2 h 16 min
- 40Keyword queries for database searchingVideo · 13:15
- 41Scispace and other tools for searching databasesVideo · 12:15
- 45
- 15
- 16
- 50
- 51
- 52
- 14
References and writing in MS Word 3 videos · 3 h 17 min
- 11
- 46
- 48
AI tools, plagiarism and AI detection 8 videos · 41 min
- 13
- 4
- 5
- 25
- 42How to write using AI without AI detection?Short · 0:53
- 47
- 18
- 17How to Write a Paper QuicklyShort · 1:12
Time management and mentorship 4 videos · 17 min
- 6Why Mentorship is Crucial in Scientific Writing?Video · 6:53
- 30Plan your time for review paperVideo · 7:05
- 43How to manage your time for focused workShort · 1:57
- 19Assignment for StudentsAssignment · 0:45
Posters, slides and sharing your research 5 videos · 1 h 59 min
- 53
- 54
- 55
- 56
- 49
No videos match. Try another word.
Links and sources
- 1Scientific Writing Zero-to-Hero playlistAll 56 free videos
- 2Scientific Writing Master ClassComplete recording, 2 h 46 min
- 3Using AI tools in Scientific WritingFree course with certificate
- 4Systematic Review and Meta-analysis coursePaid two-month program
- 5PRISMA 2020Reporting checklist and flow diagram
- 6Cochrane HandbookMethods for systematic reviews
- 7PROSPERORegister your systematic review protocol
- 8EQUATOR NetworkReporting guidelines for every study type
- 9ICMJE RecommendationsAuthorship, AI use and publishing ethics
- 10COPECommittee on Publication Ethics
- 11Think. Check. Submit.Avoid predatory journals
- 12metafor (R)Free meta-analysis package
Related guides on Codanics
- 1Using AI tools in Scientific WritingFree course with certificate
- 2Systematic Review and Meta-analysisPublish an SR and MA
- 3Conference posters with AI toolsPresent your research
- 4AI tools for students and academicsTools for research and study
- 5Prompt engineering guideGet better answers from AI
- 6CV, motivation letter, SOP and LORApply for research positions
- 7Study abroad for a PhDResearch careers abroad
- 8Marie Curie PhD positionsFunded PhDs in Europe
- 9Postdoc and MSCA fellowship guideAfter your PhD
- 10ABC of Statistics (free book)Statistics for your results section
- 11LinkedIn networking guideShare your publications
- 12Why learn bioinformaticsData skills for life scientists
More research, AI and career guides on the Codanics blog.
Key terms explained
- IMRaD
- The standard structure of a research paper: Introduction, Methods, Results and Discussion.
- Research gap
- Something important that is unknown, unclear or contradictory in the current literature.
- PICO
- Population, Intervention, Comparison, Outcome: a format for writing focused research questions.
- Boolean operators
- AND, OR and NOT, used to combine search terms in databases.
- Narrative review
- An expert summary and interpretation of a research field, without a fixed search protocol.
- Systematic review
- A review that uses a pre-defined, reproducible method to find, select and appraise all studies on a question.
- Meta-analysis
- A statistical combination of results from several studies into one pooled estimate.
- PRISMA 2020
- The reporting guideline for systematic reviews, with a 27-item checklist and a flow diagram.
- PROSPERO
- An international register for systematic review protocols with health-related outcomes.
- Risk of bias
- The chance that a study's design or conduct leads to wrong results.
- Heterogeneity (I²)
- How much study results vary beyond what chance would explain.
- Forest plot
- A chart showing each study's effect and confidence interval, and the pooled effect.
- Predatory journal
- A journal that charges fees without real peer review or editorial standards.
- Plagiarism
- Using others' words, ideas or figures, or your own published work, without proper citation.
Frequently asked questions
What is scientific writing?
Scientific writing is clear, precise and evidence-based writing that reports research so others can understand, check and build on it. It includes theses, research papers, review articles, systematic reviews, meta-analyses, proposals and conference abstracts.
What is the difference between a narrative review and a systematic review?
A narrative review is an expert summary of a field without a fixed protocol. A systematic review answers one focused question using a pre-registered, reproducible method to search, screen and appraise all relevant studies, and is reported with PRISMA 2020.
What is the difference between a systematic review and a meta-analysis?
A systematic review is the method of finding and appraising all studies on a question. A meta-analysis is the statistical step that combines the numerical results of those studies into one pooled estimate. A meta-analysis is usually done as part of a systematic review.
In which order should I write a research paper?
Most researchers prepare figures and tables first, then write the methods and results, then the discussion and introduction, and the title, abstract and keywords last, because the abstract summarises the finished paper.
How do I find a research gap?
Read 5 to 10 recent review articles in your field, note what they describe as unknown, contradictory or understudied, check that newer papers have not already answered it, and turn it into one to three specific research questions.
Can I use ChatGPT, Claude or other AI tools to write my paper?
You can use AI tools to search, summarise, brainstorm and improve language, but you must write the science yourself, check every fact and reference, follow your journal's AI policy and disclose AI use. ICMJE and COPE state that AI tools cannot be listed as authors.
Do I need to register a systematic review protocol?
It is strongly recommended. Health-related systematic reviews are commonly registered on PROSPERO before screening starts, and other fields can register on the Open Science Framework. Many journals expect a registration number.
Which software do I need for a meta-analysis?
Free options include R with the metafor package and Cochrane's RevMan. You also need a reference manager such as EndNote or Zotero and a screening tool such as Rayyan.
Is the Scientific Writing course free?
Yes. The Scientific Writing Zero-to-Hero playlist (56 videos) and the Master Class recording are free on YouTube, and the Using AI tools in Scientific Writing course is free with a certificate. The Systematic Review and Meta-analysis course is a paid program.
How do I avoid predatory journals?
Check that the journal is indexed in Scopus or Web of Science, has a real editorial board and transparent peer review and fees, and run it through the Think. Check. Submit. checklist. Be careful with unsolicited emails inviting you to publish quickly.
No questions match. Try another word.
About the author. Dr. Muhammad Aammar Tufail is the founder of Codanics and completed his PhD in a Marie Skłodowska-Curie programme. He has trained hundreds of students in scientific writing, systematic reviews and meta-analysis, and teaches research, AI, data science and bioinformatics in Urdu and Hindi.
See his research on Google Scholar and ResearchGate, and follow him on LinkedIn.
Journal policies, especially on AI use, change often. Always check your target journal’s latest author guidelines.