From now until August 31, 2026, the Ross Pendergraft Library is offering a free trial of the AI-power academic search engine, Consensus, to all Arkansas Tech University students, faculty, and staff.
What Is Consensus?
Consensus is a search engine for academic content that uses artificial intelligence to provide summaries of a given result set that may provide insight, answers, or a ‘consensus’ of highly-cited academic research on a given topic. It claims to make academic searching easier and less time-consuming for busy researchers.
How Do I Create An Account?
Tech users can create an account with Consensus using their atu.edu email. This will automatically grant them Consensus enterprise level account status from now until next year. If users have already created an account with their atu.edu email, users will see the premium icon in the bottom, left-corner of the Consensus homepage:
The current trial enables users to have unlimited Pro searches, unlimited Study Snapshots, and 50 Deep Searches a month.
What Content Does Consensus Search?
Unlike general search engines like Google, Consensus will only search academic articles and peer-reviewed studies. It uses sources like Semantic Scholar, OpenAlex, Pubmed, and its own collections licensed through other academic publishers. Some of the results include preprint versions of articles before they were peer-reviewed and published elsewhere. According to their website, users can search over 220 million peer-reviewed articles. While most results will be comprised of open access content, it may also include articles with abstracts only or restricted access. These articles can be searched by title or DOI in the library’s search engine for immediate full-text access or access through Interlibrary Loan. You can see some illustrations of their included content at their dashboard.
How Do I Use It?
Like any search engine, the website features an open search box where users can query the database. It supports both semantic searching and structured, Boolean searching. A semantic search, like a whole phrase or question, searches for synonyms and contextual meaning rather than the exact search terms. A Boolean search will give you a more precise result set featuring articles using exact keywords, and the ability to recognize operators like AND, OR, and NOT, as well as quotation marks for phrase-searching.
If you ask Consensus a “Yes” or “No” question, this will activate a meter which will synthesize results from the top 20 papers to provide an illustration of where academic consensus on your question would fall:
View their Best Practices page to get more out of your Consensus search.
What Makes Consensus Different Than Other Academic Search Engines?
Like all academic search engines, Consensus uses search filters to help users refine results. It also ranks results by relevancy to your search query. It does have a few features that make it stand out:
- The first twenty results of every search are summarized to give a broad overview of the result set.
- Consensus uses citation count, recency, and journal impact in result ranking.
- Users can use an embedded chatbot to ask questions about an individual paper.
- Paper “Snapshots” provide a quick list of a study’s methods, sample sizes, population, location, outcomes, and results.
- Advanced filters on study methods, citation count, sample size, duration, and journal ranking are available.
- Users can create a table, draft outlines, or upload other papers to analyze, summarize, or query.
Unlike the Library’s collection of databases, it cannot access our own licensed content, nor can it search across our own premium collections. It also lacks robust controlled vocabulary searching features (subject searching).
Additionally, some library databases from Ebsco and Proquest already include article summarization in their results. JSTOR users with individual accounts can also access a “Chat With A Paper” feature and view an AI summarization of the first twenty results.
Should I Trust Consensus?
While this search engine cannot hallucinate by fabricating citations (like ChatGPT), it can provide misleading summaries of real articles. Researchers should treat any AI-generated summary with the same skepticism reserved for any second-hand source of information. It can leave out important details or misinterpret results.
Students should also be made aware that copying an AI-generated summarization without attribution can constitute plagiarism. They should review the Academic Integrity policies of ATU or discuss using this content with their professor.
Interested users should review their Security and Privacy Policies before creating an account. While the product claims to not use customer data to train AI models or sell data to third-parties, it is still a private company, and individuals should carefully assess risks to their data and privacy before creating accounts.
What’s the Consensus On Consensus?
We want to know your feedback and experiences using this new tool—good or bad. As artificial intelligence changes the way we learn and conduct research, it is important to use these opportunities to be curious, be skeptical, and be responsible in our practices. If you want to give us your impressions on the tool, or if you have additional questions or concerns about using it, reach out to the library by contacting us via email at askus@atu.libanswers.com.
Prepare to be totally creeped out this Wednesday, October 29th at 6:00 P.M. in RPL 300B as we welcome a guest presentation from Amy Milliken on the spooky stories and folklore centered around several historic buildings at Arkansas Tech University.
