The eight-second attention span — a discourse object, audited

Frame

This study does not produce novel citation tracing. The work of walking the chain backward from viral references to the unverifiable institutional sourcing has already been performed, between January 2016 and February 2017, by Jon Schwabish (PolicyViz), Morton Ann Gernsbacher (UW Madison), Simon Maybin (BBC News), and Jo Craven McGinty (Wall Street Journal). Each independently identified the citation chain's terminus at Statistic Brain Research Institute and confirmed that the institutional sources Statistic Brain named — the National Center for Biotechnology Information, the U.S. National Library of Medicine, and the Associated Press — could not retrospectively verify the underlying research. Re-litigating the trace is not the framework's contribution.

What the framework can audit is the eight-second attention span as a discourse object: the propagation chain through which a content-mill aggregator's unsourced figures, recited on Marketplace by Harvard Business School historian Nancy Koehn in February 2014 and reprinted on page six of Microsoft Canada's Spring 2015 marketing-research report, became established cultural fact. SAF treats the discourse object the way KIT-3 treats any source-of-claim package: ask what the package asserts, where each assertion came from, what evidentiary weight the source warrants, and what surviving contradictions appear when the package is set against independently established evidence — including evidence within the propagating document itself.

The frame matters because the verdict cluster will lean toward demolished. A reader could fairly ask whether running SAF against a claim already debunked by four independent investigators is a fair test of the framework. The kit's discipline holds the audit accountable to running the canonical charitable-interpretation menu against each contradiction before the verdict is reached — not as a courtesy, but because charity is a structural test of whether the case is still standing. Where the strongest charitable reading of the propagated claim fails to recover it, the demolition is not rhetoric; it is the discipline producing the verdict the discipline was designed to produce. The audit's distinctive contributions — integrating the pre-Microsoft Marketplace/Koehn citation into the chain analysis, surfacing Microsoft's internal-report contradiction as a logically independent failure mode, and treating the construct conflation as separable from the citation collapse — are structural reframings, not new evidence.

Synthesis

What the audit finds.

The eight-second attention span claim does not survive structural audit on three logically independent grounds. First, the citation chain terminates at a content-mill aggregator whose listed institutional sources have explicitly disclaimed any record of the underlying research, and Microsoft Canada is not the chain's origin — the figures were already in mass-media circulation through Marketplace fifteen months before Microsoft's report. Second, "average human attention span" is not a coherent univariate construct that admits a single seconds-measurement; Microsoft's own report adopts Sohlberg and Mateer's three-part attention taxonomy (Sustained, Selective, Alternating) and uses an EEG measurement output structured as ACE (Attention, Connection, Encoding) — neither framework produces or supports the figure that propagated. Third, the same Microsoft report whose page six slide spawned the viral claim contains, in its primary EEG findings on pages twenty through twenty-two and in its executive summary on page four, data and language that contradict the slide's "dwindling" framing. The press repetition through TIME, USA Today, the Telegraph, NBC Nightly News, the New York Times opinion section, and others laundered the contradiction by quoting the slide and ignoring the data — Timothy Egan in the New York Times explicitly admitted he had not read the report. Each demolition mode survives independently of the others.

What survives charitable consideration.

A subset of the propagation apparatus survives partial charity. Microsoft Canada's slide attribution ("Source: Statistic brain") is editorially honest under the conventions of marketing-research publication. The slide does not present the figures as Microsoft's own research, and the report's actual methodology — a quantitative survey of two thousand Canadians and an EEG neuro deep-dive of one hundred and twelve Canadians, conducted in late 2014 with Brainsights as the EEG vendor — does not generate or claim the seconds-figure on page six. Charity recovers Microsoft's authorial intent: a marketing-frame opening slide editorially decoupled from the analytical findings that follow. Charity also recovers the press's structural conditions: TIME and the secondary outlets repeated the claim under deadline pressure in a media environment where major corporate research publications function as credibility heuristics. None of this charity recovers the claim itself.

What dissolves.

The institutional sourcing at the chain's terminus does not survive any charitable reading available to the discipline. NCBI is database infrastructure, not a research publisher; the U.S. National Library of Medicine has no record of the underlying research and has confirmed as much when contacted by three independent investigators; the Associated Press could not locate any article containing the figures. Statistic Brain's own About Us page declares the operators are not analysts, and the entity has not responded to multiple investigator inquiries spanning 2016 through 2017. Microsoft's internal-report contradiction is recoverable as editorial intent (dual-register publication, marketing-frame slide rather than empirical finding) but not as artefact-level consistency: the same document whose slide propagated the claim contains, on the page directly opposite a sustained-attention demographic chart, EEG indices showing Encoding scores improving for the heavy social-media and early-tech-adopter populations whose Attention scores decline. The propagated claim survives only by stripping away the rebuttal that the same document supplies.

What the discipline preserves.

The dissolution of the eight-second claim does not erase the phenomenon the claim distorts. The legitimate research surface on attention in digital environments is the longitudinal work led by Gloria Mark with Shamsi Iqbal, Mary Czerwinski, and colleagues at UC Irvine and Microsoft Research across CHI and CSCW publications from 2008 onward, synthesized in Mark's 2023 book Attention Span. Their data documents the decline of workplace screen-attention duration from approximately one hundred and fifty seconds in 2004 to seventy-five seconds in 2012 to approximately forty-seven seconds in recent measurements, replicated by independent researchers within ±3 seconds. Mark's research measures task-switching frequency in workplace digital environments, not biological attention capacity. Conflating measured screen-switching with biological capacity is itself the same category error the eight-second claim commits, and the discipline declines to recover the propagated claim by importing the legitimate finding into a frame the legitimate research does not support.

What changes verdicts.

The verdicts in this audit reverse only on evidence that has not surfaced across nearly a decade of citation tracing. Each contradiction below carries an explicit falsifiability statement specifying the conditions under which a verdict would reverse: a primary research source predating February 11, 2014 (the Marketplace/Koehn citation); a documented Statistic Brain methodology with sample size and dating; an archive at NCBI, NLM, or the Associated Press containing the figures with an authoring research team; a pre-2015 Wayback capture of Statistic Brain showing "2013: 8 seconds" rather than the post-2015 "2015: 8.25 seconds" framing; or internal Microsoft documentation establishing that the page six slide was intended as a separate empirical claim distinct from the report's EEG findings. Subsequent revisions of this study will treat substantive falsifiability material as input, surfaced through the audits feedback channel.

Method note

This study uses only public-record sources. Microsoft Canada's "Attention Spans" report (Spring 2015, Consumer Insights / Microsoft Advertising), Statistic Brain's "Attention Span Statistics" page (verified via Wayback Machine captures and via direct screenshot in Gernsbacher's Wisconsin Psych 532 packet), the Marketplace segment of February 11, 2014 (Kai Ryssdal hosting Nancy Koehn), the BBC News piece by Simon Maybin (March 10, 2017) with named expert quotes from Gemma Briggs at the Open University and Felicity Huntingford on goldfish cognition, the PolicyViz piece by Jon Schwabish (January 29, 2016), the Wall Street Journal piece by Jo Craven McGinty (February 17, 2017), and the peer-reviewed primary literature on goldfish cognition (Brown 2015, Animal Cognition) and on workplace screen-attention (Mark, Iqbal, Czerwinski et al., CHI and CSCW publications 2008 through 2018, plus Mark 2023) constitute the primary evidence corpus. Where Microsoft EEG values are quoted, the source attribution traces to the report's published pages seventeen through twenty-two as preserved in the Wisconsin packet's verbatim screenshots and in the squarespace mirror PDF reading.

The kit's brand commitment to symmetric scrutiny holds in this study by treating the propagated claim as it would treat any institutional account — with the canonical charitable-interpretation menu run against each contradiction before verdict, and with falsifiability conditions attached so the verdicts are checkable. The discipline applied here is the same the kit would apply to any claim on any subject. The audit credits Schwabish, Gernsbacher, Maybin, and McGinty as prior-art investigators whose substantive citation work this study builds on rather than duplicates; the contribution is structural reframing, not evidentiary discovery.