Systematic Literature Review Webinar: 8 Steps With Dr. Faheem Ullah

In short: A systematic literature review follows a fixed, documented method, so another researcher could repeat it and reach the same findings. In this SciSpace webinar, Dr. Faheem Ullah (University of Adelaide) teaches eight steps, from defining the need for the review to analysing and reporting the data. His advice: keep to two to six research questions, pilot them on 8–10 papers you already know, and expect reviewers to question a review with fewer than about 50 studies.

This page has the full recording, every chapter, the method step by step and the questions researchers asked live. For the written method with a full PRISMA worked example, read How to Do a Systematic Review Step by Step.

Watch the webinar

Key moments:

• 1:16 What is a systematic literature review vs a normal review?
• 5:52 The 8 steps of a systematic literature review
• 9:56 How many research questions should a systematic review have?
• 12:53 How to design a search string and choose databases
• 14:51 PRISMA screening and snowballing: from 4,634 papers to 74
• 19:30 How many papers should a systematic literature review include?
• 20:52 Data extraction form and analysis methods
• 25:00 Automating data extraction from papers with SciSpace
• 27:06 Comparing research papers critically with SciSpace
• 29:26 Creating diagrams for a literature review with AI
• 30:21 SciSpace SLR agent: from research questions to a PRISMA flow diagram
• 33:23 Checking an SLR for plagiarism and AI-written content
• 35:00 Turning a paper into slides
• 36:49 AI skills for researchers: prompt engineering and ethical AI use

What makes a review systematic

A systematic literature review is a structured, methodological way of identifying, evaluating and synthesising the research available on a specific question. Dr. Ullah gives three reasons to use one instead of a narrative review:

• Replicability. Another researcher who follows your documented method should reach the same findings.
• Less researcher bias. Every decision follows rules set in advance, so personal views have little room to shape the findings.
• A comprehensive overview. You gather all the evidence on a topic, which makes the findings more credible.

Fields set out their own guidelines. In software engineering, where Dr. Ullah works, reviews follow Kitchenham's guidelines; reporting across fields follows PRISMA 2020.

The 8 steps, as taught in the webinar

1. Determine the need for the review. Be clear why you are doing it: to find a research gap, to understand a field, or as part of a master's or PhD thesis.
2. Define your research questions. Usually two to six, specific enough to answer in depth, and agreed with a supervisor or peers. See how many research questions a systematic review should have.
3. Pilot the questions. Test them on 8–10 papers you already know are relevant, to check the literature can actually answer them.
4. Design the search string. Choose the keywords and combinations that will find the right papers in each database. PICO question examples show how to turn a question into search concepts.
5. Set inclusion and exclusion criteria. A search can return thousands of results; criteria decide which papers belong in the review.
6. Screen and select studies. Titles, then abstracts, then full texts, plus snowballing, with every count recorded for the PRISMA flow diagram.
7. Extract the data. Pull the same fields from every included study into a data extraction form.
8. Analyse and report. Synthesise the extracted data and report the findings.

“If you define the wrong questions, you end up spending three or four months on something which is not going to be of any use.” — Dr. Faheem Ullah, on why the research questions come first

Questions answered in this webinar

Do each step with a SciSpace agent

• Systematic Literature Review agent: search, screen, extract and report an audit-ready systematic review (steps 1–8).
• PICO Generator: turn your research question into PICO concepts for the search (steps 2–4).
• Title & Abstract Screening agent: screen studies against your inclusion and exclusion criteria (step 6).
• PRISMA 2020 flow diagram agent: turn your screening counts into a PRISMA 2020 flow diagram (reporting).

Related guides

Frequently asked questions

注意转载授权 · 阅读原文