I recall a few years ago our teacher librarian running a research session for students studying Politics regarding a dangerous chemical that was common in our homes, our schools and our community. That this chemical was the primary cause of excessive sweating, if inhaled it could lead to death, and prolonged exposure to its solid form could cause tissue damage. A professional looking scientific website that listed the dangers of this chemical, Dihydrogen Monoxide, provided a range of scenarios where this chemical would impair and damage us, and called for us to contact our local members of parliament so that action could be taken. It implied that the government was actively trying to prevent us from knowing of this chemical’s harm.
Dihydrogen Monoxide is more commonly known as water. The website was a parody designed to expose how susceptible we can be to misinformation, as well as exploiting a lack of scientific literacy. Our librarian’s intention in running this activity was to make us aware of the need of being critical researchers, and critical consumers of knowledge, and that research frameworks for assessing information help us find accurate and reliable information.
The ability to research effectively is an important skill across many Learning Areas, and an invaluable skill for life. Over the years students have been taught using the C.R.A.A.P framework to evaluate a source for its intended authority, accuracy and purpose. However, with the increasing use of Artificial Intelligence to summarise sources in search engines, we recognise that the previous framework while still relevant in intent, was not as successful in guiding our actions with how information is currently presented. Further, the adaptive nature of AI in search engines may actively support confirmation bias in our research, and can be a conduit for misinformation and in some cases disinformation.
In our Artificial Intelligence Policy, we refer to the SIFT method as being the preferred framework we will be using when undertaking research, and more broadly when considering sharing information. Not only is SIFT a great acronym in communicating the active search for accurate and reliable information, each of the letters represents a logical step in the research process.
SIFT stands for:
- Stop: pause before sharing or acting on information.
- Investigate: check the source's credibility.
- Find: seek better coverage or context.
- Trace: track claims to their original sources.
Stop: pause before sharing or acting on information. In a world that is competing for our attention, some sources may seek to provoke an emotional response. Rather than trusting it because of how it may support what you know, stop to consider how it makes you feel, as sometimes an emotional response can compromise our ability to think critically about information.
Investigate: check the source's credibility. What sites have been used in the generation of the AI summary? How trustworthy are these sites? Is the purpose of these sites to inform, such as news or educational sites? What can the ‘URL’ of the site tell us about the reliability of the content from the site? Does the source have expertise on this topic?
Find: seek better coverage or context. Step outside the search engine, or the AI tool. Don’t accept the first search result but look for other trusted sources to corroborate information by opening new tabs on the topic. This is sometimes referred to as lateral reading, where you open adjacent tabs to find other sources on the topic, and perhaps find information on the origin of the first source. This assists in validating claims, identifying motives and can help avoid deception. There have been items in the news recently of a lawyer submitting documents that referred to non-existent cases, due to what is sometimes called an AI Hallucination.
Trace: track claims to their original sources. If a site provides a claim or a quote, trace it back to the original context. Sometimes quotes can be taken out of context, and the headlines of an article may be sensationalised to present a particular perspective.
The SIFT framework is an important approach to help ensure that we are critical consumers of information and it is an efficient method to identify the reliability of information. Unlike the previous framework, SIFT acknowledges the emotional responses that online content often seeks to generate. By ‘stopping’ it allows us to ‘breathe’ and follow a process to discern the accuracy and reliability of the information being provided.