SciSpace PRISMA Literature Review: Amazing...Until This Happened

Key Concepts

• PRISMA (Preferred Reporting Items for Systematic Reviews and Meta-Analyses): An evidence-based minimum set of items for reporting in systematic reviews and meta-analyses.
• AI Agents: Automated software tools designed to perform complex academic tasks, such as literature searching and data extraction.
• PICO Framework: A mnemonic used in evidence-based practice to frame research questions (Population, Intervention, Comparison, Outcome).
• Boolean Search Strings: A method of combining keywords with operators (AND, OR, NOT) to refine search results in databases.
• Hallucination: A phenomenon where AI generates false or non-existent citations; mitigated here by "Verified Report" features.
• Systematic Literature Review (SLR): A rigorous, structured method of identifying, evaluating, and interpreting all available research relevant to a particular research question.

---

1. The SciSpace Systematic Literature Review Process

SciSpace provides an automated workflow for conducting an SLR. The process is designed to be interactive, requiring user input at critical junctures to ensure the research remains aligned with the user's goals.

Step-by-Step Methodology:

1. Initialization: Select the "Systematic Literature Review" tool and define the research topic (e.g., "mobile use and depression and anxiety").
2. Verification Setup: Enable the "Verified Report" feature. This is a critical step that forces the AI to cross-reference citations against real databases, significantly reducing the risk of AI hallucinations.
3. PICO Formulation: The AI generates research questions based on the topic, which the user must review and confirm. This stage is vital for setting the scope of the review.
4. Screening Criteria: The AI proposes inclusion and exclusion criteria. Users are encouraged to treat this as a learning opportunity to understand why certain studies are excluded, rather than blindly proceeding.
5. Search Strategy Configuration: Users define the search volume (up to 3,000+ papers), select databases (SciSpace, Google Scholar, PubMed, or personal Zotero libraries), and set publication date ranges. The AI converts PICO elements into Boolean search strings.
6. Execution: The AI runs the search, filters duplicates, extracts data, and generates a PRISMA flow diagram and a comprehensive markdown report.

---

2. Critical Analysis: The "Fatal Flaw" in Pricing

The author highlights a significant lack of transparency regarding the cost of these operations.

• The Issue: The platform uses a credit-based system that is not clearly mapped to the resource intensity of an SLR.
• The Experience: The author initiated a review expecting their existing subscription to cover it, only for the process to halt halfway through due to credit exhaustion.
• The Cost: Completing a full-scale SLR (searching ~2,600+ papers) required upgrading to an "Advanced" plan, costing approximately $90/month.
• The Critique: The author argues that SciSpace should provide an upfront estimate of how many credits a specific search volume will consume, as the current "opaque" system creates frustration and a "sour taste" for users.

---

3. Output Quality and Ethical Usage

Despite the pricing frustrations, the author concludes that the quality of the output is exceptional.

• Technical Output: The final report includes a structured methodology, a PRISMA flow diagram, a results section, a discussion, and a massive bibliography (e.g., 300+ verified references).
• Efficiency: The author notes that this process, which would typically take months of manual labor, was completed in approximately one hour.
• Ethical Framework: The author emphasizes that this tool is not a replacement for human scholarship. It should be used to:
  • Structure understanding.
  • Shortcut the most tedious aspects of data gathering.
  • Serve as a foundation for a researcher to perform their own manual verification and synthesis.
• Significant Statement: "This isn't about sending this off to a journal and being like, 'done.' That is not the right way and not the ethical way to use these outputs."

---

4. Synthesis and Conclusion

SciSpace’s AI agent is a powerful tool for researchers looking to automate the labor-intensive phases of a systematic literature review. By integrating verified citations and structured PRISMA reporting, it provides a high-quality, reliable output. However, users must be prepared for the high cost of entry and should approach the tool as a "research assistant" rather than an "author." The primary takeaway is that while the technology is transformative for academic productivity, the platform's pricing transparency needs significant improvement to match the professional expectations of its user base.