Lesson Specification: Module 3.4

Metadata

Learning Objectives

By the end of this lesson, students will be able to:

1. Describe the full PRISMA 2020 workflow from protocol registration to reporting.
2. Map each phase of the SLR process to the specific Nexus tool that automates it.
3. Construct a PRISMA flow diagram using real numbers from their own search.
4. Evaluate the difference between a narrative review and a systematic review, and when each is appropriate.

Key Concepts

Concept 1: What Makes a Review "Systematic"?

• Narrative Review: An author-driven survey of the literature. The author picks what to include based on their judgment. Susceptible to cherry-picking and confirmation bias.
• Systematic Review: A protocol-driven, reproducible search of the literature. The inclusion/exclusion criteria are defined before the search begins. Every step is documented so another researcher could replicate your exact results.
• Why it matters: A systematic review is considered the highest level of evidence in evidence-based medicine and is increasingly expected in engineering, education, and social sciences.

Concept 2: The PRISMA 2020 Framework

The PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses) framework defines the standard workflow:

1. Protocol Registration: Pre-register your search strategy (e.g., on PROSPERO) to prevent bias.
2. Identification: Run searches across multiple databases. Report how many records each database returned.
3. Screening: Apply inclusion/exclusion criteria. First by title/abstract, then by full text.
4. Included: The final set of papers that passed all filters.
5. Reporting: Generate the PRISMA flow diagram showing exact numbers at each stage.

Concept 3: The Nexus Toolkit as an SLR Engine

This is the capstone lesson that ties together everything the student has learned:

• Identification: scholar-search-kit (multi-source search) + scholar-search import (external databases)
• Deduplication: scholar-search dedup (remove duplicates across sources)
• Full-text retrieval: scholar-pdf-kit (download open-access PDFs)
• Analysis: scholar-graph-kit (citation networks, snowballing) + scholar-rag-kit (semantic querying)
• Synthesis: The extraction template and synthesis matrix from Module 3.1

Concept 4: Common Pitfalls

• Moving the goalposts: Changing your inclusion criteria mid-search to include papers you "like" undermines the entire methodology.
• Publication bias: Positive results get published more often. Your SLR should acknowledge this.
• Scope creep: An SLR with 5,000 included papers is unmanageable. Tight, precise research questions (Module 2.1) prevent this.

Suggested Hooks & Motivation

• Pain Point: "Your thesis committee asks: 'How do we know you didn't cherry-pick these papers to support your argument?' You have no answer. A systematic review IS the answer — it proves your literature search was objective, reproducible, and exhaustive."
• Wow Moment: "Everything we've learned in this course — search, dedup, PDFs, graphs, RAG — clicks together in this one lesson. The SLR is the final boss, and the Nexus toolkit is your weapon."
• Real-World Scenario: "In fields like medicine, a systematic review can literally change clinical practice guidelines. In CS and engineering, SLRs are increasingly required for top-tier journals."

Cross-References

• Builds on: Every module in Modules 2 and 3.
• Leads to: Module 4.1 (Writing — now you have the content to write about).

Sources & Further Reading

1. Page, M. J., et al. (2021). The PRISMA 2020 statement. BMJ.
2. Kitchenham, B. (2004). Procedures for Performing Systematic Reviews. Keele University Technical Report.