Three tools to help you navigate and review IS literature

LitBaskets, PermuSearch & CoLRev

Before we start

• Create a GitHub account and log in
• Start the demo
• The slides are available online:

Our background

Philosophy: Reliability, efficiency, and contribution

Reliability

• Transparency and openness of data, and open-source code
• Validation of data and code
• Efficient undo operations

Efficiency

• Tool testing and innovating
• Team expansion, involving colleagues, student assistants, and crowds
• Data reuse, e.g., updating prior reviews, importing samples from other review papers

Contribution

• Tools must build on a nuanced understanding of review goals and types
• Knowledge contributions primarily depend on the ingenuity of researchers (with some help of tools)

Literature reviews with CoLRev

• Based on Git for data versioning and collaboration
• An open and extensible environment supporting different types of reviews
• Covers all steps of the process

The CoLRev tutorial

• Form small groups (2-3 people) and work together
• Go to the codespaces, read the worksheet and enter the commands
• You can enter colrev status at all times
• Consult with the documentation when necessary

How to get involved

Open questions

Use AI in the search to reduce time

• ChatGPT can be useful for exploratory searches (see Wang et al.)
• See Wagner, Lukyanenko and Paré 2022 and the follow-up paper (currently under review)
• A CoLRev package for genAI is currently under development
• If the challenge is more significant in the prescreen - consider the active-ML tool ASReview

Can not find the right information

• Use litbaskets for journal scope and Permusearch for search terms
• Iterate and extend search based on keywords and terms from the papers
• Consider guidelines of Boell and Cecez-Kecmanovic 2014 (Appendix A)
• Once you have a working strategy: report and publish it (e.g., at searchRxiv)

Finding high-quality literature outside of IS

• Consider a backward search with a number-of-citations threshold