The information available from the below resources are summaries (aka includes information from of a lot of studies that may or may not include expert option) or written from a consumer health perspective. The below may be published peer reviewed articles (eg. a Cochrane systematic review), or a website with background information about a condition.
The below databases are multidisciplinary. This means they contain articles from many different disciplines such as medicine, nursing, social sciences, history, education and political science. If you're looking for articles for a question that is multidisciplinary - start your search here!
You can also see our full list of Electronic Resources.
Full-text access to articles from 1,700 Elsevier journals covering the subject areas of science, social sciences, arts, and humanities.
Look for the “Check for full text @ X” link next to each result. Use that link to get the full text via the Library’s subscriptions.
Link your Google account to StFX, so the “Check for full text @ X” links will always appear – even if you don’t start your search from this page. Go to our customized settings screen and click Save. If you’re logged into a Google account, it will remember that setting from now on.
There are a lot of new information discovery tools available that are enhanced with AI and LLMs. There seem to be 3 main categories emerging of 'AI Powered Search' tools:
1) AI Search Engines
These are the chatboxes that have the ability to search the open web and provide references the the text generated. Some prompting may be required. eg. ChatGPT with search, Perplexity, MS CoPilot with search
2) AI Academic Search Engines.
These are grounded in scholarly sources only and do not require prompting. They can be add on to traditional library resources eg. Web of Science Smart Search or direct to consumer products eg. Elicit or Consensus
3) Deep Search
These are semi-automated research assistants/ 'agentic' search. These products can be grounded in scholarly sources only or the open web. The search will take a long time (about 5-10min). eg. Undermind.ai, ChatGPT Deep Search or Perplexity Deep Search.
See this webinar by Aaron Tay (Oct 9, 2025) for more information Implications of Ai powered academic search.
Most of these tools use a technology called Retrieval Augmented Generation (RAG).
If you are going to be exploring using these tools, watch the video below by IBM explaining what RAG is.
There are 3 main changes from traditional search to be aware of:
1) The search technology
The search starts with a natural language question or phrase. From there the tool will either:
Most AI Academic Search Engines seem to use both LLM- Generated Boolean and Semantic Search.
2) Opaque LLM Prompts and Relevancy Rankings
There are a lot of default system prompts and programs running between what a user types in and the information retrieved. There are very complicated relevance ranking algorithms running on the backend to decide what the top "most relevant" results to use.
3) Output
Where traditional tools display a list of citations for you to pick from, these tools display answers with a summary and citations or a full structured report.
There are also some components of these tools that you need to focus on critically evaluating if you plan to use them.
1) Accuracy of the summary of the answer (consider: do you know enough to know if it's true?)
2) Quality of the references (are they the best out there?)
3) Citation faithfulness or citation precision (eg. does the information cited supports the specific claim or information it’s being used in the summary? or was it a general statement from an introduction? or was it actually a secondary source who made this claim?)
4) Consistency of results/ findings
To see more about how AI has wrong summaries even when it is citing information sources, see this toolkit developed by the BBC.
For more information on evaluating AI Academic Search Engines in general see this checklist by UofT Libraries
AI Powered Search Tools:
Some Harms Associated with AI Use
As there are with the use of AI in general. There are also very real and serious ethical and environmental concerns around the development, use, and sustainability of these tools. See this guide by Rebecca Sweetman for more information.
See this guide by UofT Libraries for some help evaluating these tools.
See this guide by McGill Libraries for help evaluating AI enhanced resources.
