Discussion Thread: AI Research Tools and Proper Use AI Research Tools
Serenity3203
Chapter8forHtdylr.pptx
PowerPoint 8 Strategising: comparing systematic, realist and narrative reviews
Chapter 8 from
How to do your literature review
This is PowerPoint 8 for Chapter 8 …
We will look at
Linear or recursive approaches to literature review?
Realist synthesis
Revising the research question
The presentation of the review
All of this is discussed in much more detail in Chapter 8 of How to Do Your Literature Review
Should an approach to literature review be linear or recursive?
The systematic review takes a linear approach, which has to be laid out at the outset and must be followed.
Narrative reviews take a recursive approach, assuming, for example, that the question may be revised after initial scoping.
Linear approach
Recursive approach
Revised question
Lit review
Question(s)
Literature review
Question
Scoping and mindmapping
Search
Search
The best of both worlds? Combining elements of the systematic literature review with the narrative review
Propose a prima facie question
Make an initial scoping of the literature with some key references
On the basis of your scoping, draw a mindmap, and, if necessary, develop this into a storyboard
In light of the mindmap, revise the prima facie question if necessary
Specify the search strategy for the full review:
Specify databases and AI tools to be used
Specify keywords to be used
Specify inclusion/exclusion criteria
Consider the broad structure of the review. In other words, think about whether it would be best for your review analysis to be:
Funnelled
Thematic
Chronological
Comparison-based
Mixed
Analyse and synthesize according to the structure
Report
Pawson’s realist synthesis
Rather than asking simply ‘What works?’ realist literature synthesis seeks to unpack the mechanisms through which complex interventions work (or why they fail) in particular contexts.
This involves:
A background search to get a feel for the literature (this is the scoping that we have looked at).
Progressive focusing to identify the thinking and theories around the topic.
A search for empirical evidence from a wide range of sources which use a variety of research strategies.
Pawson, R., Greenhalgh, T., Harvey, G., & Walshe, K. (2005). Realist review—A new method of systematic review designed for complex policy interventions. Journal of Health Services Research & Policy, 10(1_suppl), 21–34. https://doi.org/10.1258/1355819054308530
Main differences between systematic, realist and narrative reviews
| Systematic | Realist | Narrative | |
| Design | Linear and laid out in protocol to be exactly followed | Recursive, depending on findings from scoping | Recursive, depending on findings from scoping |
| Focus and structure | Evaluative | Evaluative | Any one of several, including funnelled, thematic, chronological, comparison-based, mixed |
| Scoping processes | If any, defined at outset and clearly specified | Flexible and evolving | Flexible and evolving |
| Question | Set at outset | Set at outset and revised | Set at outset – a prima facie question – and revised |
| Search processes | Via databases using keywords | Via databases using keywords, snowballing | Via databases using keywords, snowballing, personal knowledge, AI tools |
| Attitude to data | Extraction | Extraction and interpretation | Interpretation |
| Kind of study preferred | Usually RCTs | Any | Any |
Should the research question be an anchor or a compass?
In a systematic review, the research question is unchanging
In a narrative or realist review, it may change in light of initial scoping.
Should the research question be an anchor, holding the study firm, or a compass, helping to set the direction?
How should a review be presented?
| Chapter or Heading | Subheadings | Contents |
| Abstract | Summary | |
| Introduction | Prima facie question Rationale for the review | Why is this an interesting area to research? |
| Scoping | Scoping Mindmapping Revised question | Key references from expert tutors, from search engines, AI tools, and snowballing key references How key references were used to help construct a mindmap and storyboard Whether the prima facie question was revised on the basis of the scoping and mindmapping |
| Search methods | Databases sourced | For example, PubMed, Scopus, Google Scholar |
| AI tools used | For example, Elicit.com, Connected Papers | |
| Search terms employed | Keywords or key phrases | |
| Inclusion/exclusion decisions | How judgements were made about including and excluding data from sources | |
| Structure of the review | Funnelled, thematic, chronological, comparison-based, or mixed | |
| Findings and discussion | Sections, headings or chapters | Main headings or chapters based on themes/points from the structure |
| Conclusion | Digest of main points Challenges Strengths and weaknesses |
Summary
Complexity is an abiding characteristic of applied social science. In education, healthcare and social work, for example, valid conclusions about complex situations must be embedded in a broad range of contextual information.
Whereas the systematic review promotes the importance of a linear approach to the review process, other reviews value recursiveness – going back and revising questions and methods in light of accumulating knowledge.
Realist synthesis sets great store on focusing progressively closely on a question or topic to identify the thinking and theories around the topic prior to a thorough search of the literature.
Realist synthesis values a broader range of evidence than that which is usually acceptable to systematic review.
Both systematic and realist types of reviews direct themselves to, and are most suitable for, evaluation-based questions. A broader range of questions and topics are amenable to narrative review.
Different kinds of review are not in opposition to one another but may be complementary. We should be able to mix and match the best elements of each.
Activity
Think of a general topic or devise a general question in your area of interest and write it in the shaded box on the bottom left of the table. (The one above is an example.)
Do a database search on your topic or question using a specific library database or a generic one such as Google Scholar, or an AI tool such as Elicit.com.
On the basis of your thinking and the first few ‘finds’ from your search, revise your first question so that it becomes sharper. Remember: the best questions are clear, focused and answerable.
| Start-up question/topic | Do a preliminary search | Revised question/topic |
| Narcissism in modern politics | | How does narcissism in modern politics impact political decision-making and governance, and what are the potential consequences for democratic societies? |
| Your own start-up question in a topic of interest to you | |
Example
image1.jpeg
image2.jpeg
image22.png
image23.svg
.MsftOfcThm_Text1_Stroke_v2 { stroke:#FFFFFF; }
image3.jpeg
image20.png
image21.svg
.MsftOfcThm_Text1_Stroke_v2 { stroke:#FFFFFF; }
image24.jpeg
image25.png
image26.svg
image27.png
image28.svg
