Skip to content

Misinformation and Data Literacy

Speaker: Jevin West
Event: MisinfoCon DC at UW CIP
Year: 2026
Video ID: i5gOQJGOJ-Y

TL;DR

Data and numbers are surprisingly effective misinformation tools because they convey false authority and precision—people trust numbers even more than words. West illustrates how charts can deliberately mislead through axis manipulation, bin-width distortion, and selective framing; these occur regularly in mainstream media and academic work. He presents educational interventions (the "Calling Bullshit" course and its K-12 variants) designed to teach critical reasoning about data, alongside emerging threats (deepfakes, voice synthesis) that compound the credibility crisis. The core insight: we must teach students to question data critically, the same way humanists question symbolism—not as technical execution but as interpretive work.

Key claims

Numbers carry unearned authority: People accept numerical claims more readily than verbal ones, even when the underlying claim is identical. A vague statement like "countries give us their worst people" becomes persuasive when framed numerically: "2,139 DACA recipients were convicted or accused of crimes"—but omitting context (0.3% vs. 8.6% conviction rate for US citizens) reverses the story entirely.

Chart manipulation is ubiquitous: Even high-quality publications (Wall Street Journal, New York Times, Washington Post) publish misleading visualizations. A carbon-emission chart with inconsistent x-axis increments (30-year jumps for historical data, 1-year increments for recent data) creates a false impression of "leveling off." Simply re-plotting the same data with consistent axes tells a different story.

STEM students lack critical reasoning about data: Technical competence (finding Jacobians, replicating algorithms) is common; critical questioning of numerical claims—the kind taught in humanities via textual analysis—is rare. This gap is dangerous in a data-saturated world.

Data literacy education is scaling: The "Calling Bullshit" course (co-taught with Carl Bergstrom) has been adopted by 70+ universities globally and adapted for K-12 through partnerships with state education departments. Media literacy requirements are now mandated in some state curricula (e.g., Washington State).

Synthetic media (deepfakes, voice synthesis) represent an escalating threat: The combination of image-based (photoshopping, "boyshopping" from UW's own research) and audio-based (voice synthesis) manipulation creates a new landscape where video and audio evidence can be fabricated convincingly. Public literacy around these tools lags far behind their sophistication.

Misinformation about science poses particular risks: False health claims (vaccines causing "shaken baby syndrome," Ayurvedic pseudo-cures, fluoridation myths) propagate via search engines and recommendation algorithms. West's son received flat-earth recommendations immediately after searching for International Space Station content.

The asymmetry principle: Creating misinformation requires far less energy than refuting it—an orders-of-magnitude difference. This asymmetry is structural and demands preventive literacy rather than post-hoc correction.

Mechanisms discussed

Visual manipulation tactics: - Inconsistent axis scaling (logarithmic vs. linear, jumbled increments) - Arbitrary bin-width selection in histograms (changing the bins changes the story) - Cropped y-axes (amplifying small differences) - Color and design choices that bias interpretation

Data misrepresentation: - Selective statistics (reporting one metric while omitting context) - Decontextualization (absolute numbers without reference classes) - False precision (citing decimals where uncertainty dominates)

Education-based interventions: - Teaching students to reconstruct data from visualizations - Emphasizing the constructive/interpretive nature of data storytelling - Leveraging momentum in state legislatures for media-literacy mandates - Partnering with K-12 educators and high-school teachers - Creating both digital and physical curriculum materials

Connections

Notes

This talk is valuable for its reframing of the misinformation problem as fundamentally pedagogical rather than technical. Most research focuses on detection algorithms or platform design; West argues the leverage point is population-level critical reasoning. His examples (DACA statistics, carbon-emission charts, tax-policy visualizations) are concrete and reproducible—students can literally verify the manipulation themselves.

The observation that STEM-trained students excel at technical tasks but lack humanistic critical reasoning is particularly incisive. Data literacy requires the same interpretive skepticism applied to novels or historical texts, not just technical fluency.

The "Calling Bullshit" course's adoption at scale (70+ universities, state-mandated K-12 curricula) suggests real institutional momentum around this intervention. The Knight Foundation and state legislatures are investing in this approach, making it a documented trend, not an isolated effort.

The asymmetry principle West closes with—that refutation requires orders of magnitude more energy than misinformation—is a key insight adopted across the literature. It explains why fact-checking, while necessary, cannot be the primary defense.