Get the most out of Consensus, an AI academic search tool

Scholarly sources made simple

Student working and using a laptopSMU Libraries is currently trialing Consensus.app, an AI tool that searches across about 200 million academic papers. Consensus is a search engine not a chatbot, so it only gives you real, academic papers in the results unlike other AI tools you may be familiar with (ChatGPT, I’m looking at you). Here are some tips to get the most out of your searches.

Register with your SMU email, and make sure your profile is connected to SMU.

This will get you access to the upgraded version, with unlimited Pro searches and 50 Deep Searches a month. You will also have one-click access to articles available through SMU Libraries.

Consider how your prompt impacts your results.

Like most AI tools, you get better results when you analyze the output and change your next prompt to better fit what you want. A broad topic in the prompt will retrieve a range of subtopics, good for getting your feet wet but maybe not as focused as what you need. Yes and no questions will generate a graphic that shows levels of agreement across the literature. You can use traditional Boolean search strings with quotation marks. You can search for known authors and titles. Consensus will offer similar articles to the one requested, but it can be a bit of a wild card as reasons why something is relevant vary widely.

Use the extra features.

The filters go a long way to communicate exactly what you need. For example, filtering for discipline works well whereas prompting it for the discipline does not. It also offers some interesting filters not available in other search tools, like journal rank and methodology.

Citation tracing is a very useful search strategy, and you can use Consensus to find an article’s cited and citing sources. Click into the article, look for the details button, and then references and citations.

Use it as a discovery tool, not a decision-making tool.

Not every result Consensus gives you will be specifically relevant to what you need. Ultimately, humans are better at determining what is relevant to their research than the AI tool. I have never generated a set of results in Consensus and thought that everything it returned was a homerun, even if most were in the ballpark. There are things Consensus might be missing. The summaries are useful for scanning and understanding the general layout of an area of research, but ultimately sources written by humans need to be your reference point for your citations.

Use in conjunction with other academic search tools.

There is a difference between what you need and the question you were able to ask. Sometimes, you need to scan across literature to discover something you wouldn’t have known to ask for. For literature reviews, you need to cast a wider net and go through the literature yourself. Those conducting formal systematic reviews are cautioned that using an LLM can have methodological problems.

While Consensus covers a range of disciplines, it does not search for scholarly books or other kinds of authoritative information not found in scholarly publications.

We want to hear from you!

If you have opinions about Consensus, we want to hear them! Your feedback on trial resources helps us to make informed decisions about our subscriptions.

 

This post was written by Megan Heuer, director of educational initiatives at SMU Libraries. Megan collaborates with students and faculty on their research, equipping them with contextual knowledge about artificial intelligence use in academics.