Four Moves¶

Strategy · All strategies
Evidence · 3 claims (2 for, 1 mixed) · 6 studies (3 causal, 1 review, 1 associational, 1 theoretical), q2–q3 · 1 of 6 report an effect size · 1 claim rests on one study

Description¶

The Four Moves — popularized by Mike Caulfield as SIFT — are a lightweight routine for evaluating the credibility of online information: Stop (notice your emotional reaction and resist acting on first impressions), Investigate the source (read laterally, leaving the page to learn what others say about it), Find better coverage (look for trusted reporting on the claim itself rather than judging the site in isolation), and Trace claims, quotes, and media to the original context (find the primary study, transcript, or image). The routine is carried out in seconds to minutes using browser tabs and search, mirroring how professional fact-checkers actually work. It replaces the common classroom habit of judging a page by its own surface features — design, domain suffix, "About" page — with the practice of checking what the rest of the web says.

Design Implications¶

The Four Moves work because they shift evaluation from vertical reading (staying on the page and analyzing its features) to lateral reading (leaving the page to see what independent sources say), which is the observable behavior that distinguishes professional fact-checkers from historians, PhD students, and college students in evaluation tasks Lateral reading outperforms on-page analysis for evaluating digital sources. [+S]. Because lateral reading relies on the collective judgment of the web rather than the learner's own domain expertise, it works even when the learner knows nothing about the topic — a decisive advantage over content-based evaluation for novices. Explicit instruction in these moves measurably improves students' evaluation of search results and social media content Civic online reasoning instruction improves students' evaluation of digital sources. [+S].

Context¶

Requirements¶

• Live internet access and basic search fluency — the moves are performed with the web, not from memory
• A short list of trusted reference points (e.g., established newsrooms, Snopes, PolitiFact, AP Fact Check, Wikipedia) so "find better coverage" has somewhere to land
• Authentic practice with real, current web content; evaluation skill does not transfer from lectures about misinformation alone
• Brief modeling of the moves (Think-Aloud demonstration of a fact-checker leaving a page within seconds)
• Access to online information and the ability to open parallel tabs and searches
• Basic search skills: querying a source's name plus "wikipedia," using reverse image search, finding original publication dates
• Willingness to abandon a source quickly when it fails investigation — the framework explicitly optimizes for not reading dubious material deeply
• Minimal prior knowledge of specific sources; the moves themselves supply the evaluative leverage Prior knowledge of individual sources is not required for lateral reading to improve evaluation. [~W]

Constraints¶

• Generic source-evaluation checklists (CRAAP, RADCAB) that keep students on the page produce no measurable improvement in evaluation skill [-M] — the Four Moves must be taught as actions in new tabs, not as criteria to apply to the page itself
• Novices default to vertical reading even when a page is professionally designed and deceptive [-S]; without explicit training, design polish and a .org domain reliably mislead
• Emotionally charged content that confirms prior beliefs short-circuits the Stop move; motivated reasoning can override the routine precisely when it is most needed [~M]
• In offline or locked-down environments where learners cannot open new tabs or search, moves 2–4 collapse and the routine degenerates into the checklist behavior it was designed to replace [-M]
• Classroom implementations that require exhaustive investigation of every source lose the speed that makes the routine usable in real information environments [~W]
• Ineffective when learners remain on the source page and apply the moves as an internal checklist — the benefit comes from external corroboration, not from re-reading Lateral reading outperforms vertical reading for source evaluation. [-S]
• Requires reliable third-party infrastructure (Wikipedia, established news outlets, fact-checking sites); in information environments lacking such anchors, "find better coverage" degrades into circular sourcing
• The "stop" habit is the least trained and most easily dropped move; without it, learners fact-check only claims they already doubt, confirming rather than correcting bias [~W]
• Time-pressured learners default to shallow heuristics (design quality, domain suffix) when the moves are not practiced to automaticity [-M]

Implementation Variability¶

• Full SIFT lesson sequence: Caulfield's open textbook provides scripted activities for each move, including reverse image search and Wikipedia "W:PT" checks
• Embedded practice: short recurring warm-ups in which students apply the moves to one live story per session, rather than a standalone unit
• Domain adaptation: the same moves apply to scholarly sources (trace a news claim back to the actual study) and to images/video (reverse image search, original upload tracing)
• Scaled-down version: teaching only Stop + Investigate as a two-move minimum routine for younger learners
• Taught as full lessons (Caulfield's open textbook), as short video micro-lessons (Stanford History Education Group's Civic Online Reasoning curriculum), or as embedded prompts in a research assignment
• Can be compressed to "SIFT" as a memorable acronym for quick application, or expanded with domain-specific moves (e.g., tracing scientific claims to journal articles)
• Works as a whole-class Case-Based Learning activity using live, real-world claims rather than sanitized examples

Target Learners¶

• Secondary and postsecondary students, who show large gaps between confidence and actual evaluation skill Civic online reasoning instruction improves students' evaluation of digital sources. [+S]
• Adult
• Undergraduate and secondary students conducting online research, who default to vertical reading and design-based judgments Lateral reading outperforms vertical reading for source evaluation. [+S]
• Less effective for learners who already habitually read laterally; instruction adds little for professional fact-checkers [~W]

Claims¶

• Metacognitive self-monitoring: recognizing when emotional arousal is driving acceptance of a claim Self-monitoring improves self-regulation. [+M]