Why sincere people are certain

The Bigfoot Field Researchers Organization's geographic database, the same catalog Floe Foxon regressed against black bear populations, listed 732 reports for Washington, 471 for California, 348 for Florida, 335 for Ohio and 305 for Illinois when it was checked for this entry. DOCUMENTED Those counts rise as reports are added, and Foxon used only the 2006 subset of them.

What the database does not record is anything about the people. It holds no measure of how many others a witness told, whether they expected to be believed, what they wanted, or whether they ever spoke of it again. Claims about the typical reporter get made constantly in this subject, in both directions, and this entry makes none, because no study measuring one was found for it. Hoaxes are real and this site keeps a file on them. No published estimate of what fraction of reports are hoaxes was found either, so no fraction is offered here.

The question this entry can take up is narrower and answerable: what is actually known about how an honest, sober, unmotivated person can be wrong and certain at the same time.

The instrument is worst exactly when the reports happen

The organization collecting the reports says the problem out loud. The BFRO's classification page states that the system "rates the circumstantial potential for misinterpretation, not the credibility of the witness or how interesting the report is," and reserves Class B for "incidents where a possible sasquatch was observed at a great distance or in poor lighting conditions and incidents in any other circumstance that did not afford a clear view of the subject." Class A is for "clear sightings in circumstances where misinterpretation or misidentification of other animals can be ruled out with greater confidence." DOCUMENTED What follows applies to any report made at distance, in bad light, or briefly. This entry has not verified a published breakdown of how many reports were made under those conditions, and does not assert one.

Michael Kalloniatis and Charles Luu's Webvision chapter on light and dark adaptation puts the cone mechanism in charge of vision down to about 0.03 cd/m². Below that, "the rod mechanism comes into play," and "when the rod mechanism takes over, coloured test spots appear colourless, as only the cone pathways encode colour." The handoff is slow. Rod sensitivity "improves considerably after 5-10 minutes in the dark," and the threshold curve reaches its minimum "at about 10⁻⁵ cd/m² after about forty minutes in the dark." A person who has just stepped out of a lit kitchen or a truck cab is, for several minutes, working with an instrument well short of what it will become.

The rods are also absent from the exact patch of retina people aim at things. Dale Purves and colleagues, in Neuroscience, state that "the central 300 µm of the fovea, called the foveola, is totally rod-free," that "just 6° eccentric to the line of sight, acuity is reduced by 75%," and that "the threshold for detecting a light stimulus is lower outside the region of central vision." Their own illustration of the consequence: "it is easier to see a dim object (such as a faint star) by looking away from it." DOCUMENTED Staring directly at something dim in the dark aims the least sensitive part of the eye at it. Detail has to come from a region that has effectively switched off, and shape from a region that never had much resolution.

Distance and light have been measured together for exactly this purpose. Willem Wagenaar and Juliette Van der Schrier showed three-quarter target portraits for twelve seconds, followed immediately by a six-portrait lineup, across seven distances from 3 to 40 meters and nine illumination levels from 0.3 to 3000 lux. Their result was a rule of thumb they named the Rule of Fifteen: even in ideal conditions, the diagnostic value they set as desirable, 15, is reached only at no more than 15 meters and no less than 15 lux. DOCUMENTED

T. J. Nyman and colleagues later randomized 178 participants into three lighting conditions, 300, 10 and 0.7 lux, and showed each of them eight targets at eight distances from 6 to 20 meters, every target followed by an eight-person simultaneous target-present lineup where chance alone yields .125. At 20 meters the rate of correct identifications was .53 at 300 lux, .41 at 10 lux and .11 at 0.7 lux. DOCUMENTED The authors state their own limit in the same sentence as their conclusion: "the generalizability of our findings to overall eyewitness accuracy is limited by the exclusion of target absent line-ups, yet our findings show that reliable and correct target present identifications are very unlikely following observations made in low lighting (0.7 lx) at 20 m."

Both experiments used human faces, and faces have dedicated machinery behind them. The Liu study described in the next section found a region of the right fusiform gyrus that fires specifically for faces, real and illusory. No comparable region is known to work on behalf of an unfamiliar upright animal shape. A report of such a shape, at range, in worse light, asks more of the visual system than either experiment did. Exactly how much more has not been measured.

Reported heights are the number this entry can support least, and it will not repair them. No study was found measuring how accurately people estimate the height of an unfamiliar figure at an unmeasured range with no scale object beside it. Absent one, a height in a report is best read as an impression, and no correction factor is offered here, because none has been established.

Expectation supplies what the eye leaves out

Jiangang Liu, Jun Li, Lu Feng, Ling Li, Jie Tian and Kang Lee showed participants pure-noise images and told them 50 percent contained either faces or letters. Participants "reported seeing faces or letters illusorily 34% and 38% of the time, respectively," in images that contained neither, and the right fusiform face area responded specifically when they saw faces rather than letters. Their conclusion: "human face processing has a strong top-down component whereby sensory input with even the slightest suggestion of a face can result in the interpretation of a face." DOCUMENTED

That is pareidolia measured in a scanner rather than argued about on a forum. The 34 percent belongs to that procedure, with that instruction, on that stimulus. Treating it as a general rate at which people see faces in the woods would be a misuse of it, and this entry does not use it that way. What transfers is the direction: expectation supplied by instruction changed what people reported seeing, and it changed the brain response too.

The opposite error is equally well documented and gets far less use in these arguments. Daniel Simons and Christopher Chabris had 192 observers count basketball passes in a 75-second video. Half the videos sent a person in a gorilla costume through the scene and half sent a woman with an umbrella; either way the unexpected event lasted 5 seconds. Across all 192 observers, "54% noticed the unexpected event and 46% failed to notice." The gorilla was noticed by 44 percent against the umbrella woman's 65 percent, and 67 percent noticed in the opaque videos against 42 percent in the digitally superimposed transparent ones. DOCUMENTED That result presses hardest on the confident skeptic. "I would have known if it were a bear" is the intuition the experiment breaks, and it breaks it for the reader as thoroughly as for the witness.

The courts learned this at somebody else's expense

The Innocence Project's DNA exoneration page reports 375 DNA exonerees, 69 percent of whose cases involved eyewitness misidentification, 5,284 years served between them at an average of 14 each, and that "180 of the DNA exonerees (50%) had the real perpetrator(s) identified in their cases," with the page counting 165 actual assailants identified. DOCUMENTED The organization's separate eyewitness page reports that "more than 60% of our clients were wrongfully convicted based on eyewitness misidentification," which runs on a different and larger denominator than the DNA-exoneration set. The two numbers are not the same statistic and should not be quoted as one.

What that record establishes is that misidentification is common in this set of cases. It does not by itself establish the honesty, the certainty, or the state of mind of the witnesses involved, and this entry does not claim to know those.

What certainty is worth is live and unsettled. John Wixted and Gary Wells argued in 2017 that under pristine conditions, meaning "initial, uncontaminated memory tests using fair lineups, with no lineup administrator influence, and immediate confidence statement," studies "have consistently shown that, for adults, (a) confidence and accuracy are strongly related and (b) high-confidence identifications are remarkably accurate." They stated the limit in the same abstract: where those conditions do not hold, "the accuracy of high-confidence ID is seriously compromised." CONTESTED Kimberley Wade, Robert Nash and Stephen Lindsay replied in 2018, in a commentary on a related paper by Wixted, Mickes and Fisher, that they accepted the thrust of the argument while objecting that near-pristine conditions are treated as though they were normal, when psychological science does not yet understand how often real interviews depart from best practice and witnesses are frequently exposed to influences before any interview happens. Shari Berkowitz, Brandon Garrett, Kimberly Fenn and Elizabeth Loftus argued against over-relying on eyewitness confidence in a 2020 paper drawing on exoneration data.

That dispute matters here, because a sasquatch report meets the pristine standard in no respect. There is no lineup, no administrator, and usually no confidence statement recorded before the witness has told somebody and started explaining it to themselves.

Every retelling is a rewrite

Karim Nader, Glenn Schafe and Joseph LeDoux showed in 2000 that "consolidated fear memories, when reactivated during retrieval, return to a labile state in which infusion of anisomycin shortly after memory reactivation produces amnesia on later tests." Two controls carry the argument. "The same treatment with anisomycin, in the absence of memory reactivation, left memory intact," and "delay of the infusion until six hours after memory reactivation produced no amnesia." DOCUMENTED Retrieval opens the file, and it stays open only briefly. That is rat amygdala and a conditioned fear response, not human autobiographical memory, and the entry does not assert the two are the same process.

For humans, the work is behavioral. Elizabeth Loftus's 2005 review of thirty years of misinformation research describes the effect as "the impairment in memory for the past that arises after exposure to misleading information," and closes on newer work into "the processes by which people come to believe falsely that they experienced rich complex events that never, in fact, occurred." Alan Scoboria and six colleagues then recoded 423 participants from eight published implantation studies under a single scheme: 30.4 percent of cases were classified as false memories, another 23 percent as having accepted the suggested event to some degree, and "when the suggestion included self-relevant information, an imagination procedure, and was not accompanied by a photo depicting the event, the memory formation rate was 46.1%." DOCUMENTED Those were memories of events that had never happened, held by people who had been through nothing at all.

Then the account meets other people. Henry Roediger, Michelle Meade and Erik Bergman sat a participant beside a confederate to view six household scenes for either 15 or 60 seconds each, then had the two take turns recalling six items apiece, with the confederate occasionally naming items that were not in the scene. Participants later recalled the confederate's inventions as their own more often than controls did, and the effect was larger when the scene had been shown for 15 seconds rather than 60, and larger when the invented item fit the scene, a toaster in a kitchen, than when it did not, oven mitts. DOCUMENTED A brief encounter followed by group discussion has the shape that experiment flags as most vulnerable. The tested range was 15 seconds against 60, so anything shorter than 15 is extrapolation, and this entry does not extend the curve past what was measured.

Jamie Lewis and Andrew Bartlett analyzed 166 semi-structured interviews with people involved in Bigfoot research and describe a culture in which witness accounts are the primary form of evidence and speakers work to present themselves as credible interpreters rather than credulous believers, including in how they handle field hoaxes. Whether repeated telling to a sympathetic audience makes an account longer, more coherent or more confident over time is a plausible extension of the social contagion work. It is an extension. No study testing it against sasquatch accounts was found for this entry.

The three-in-the-morning reports are their own category

Some encounter accounts are set in a bedroom. A presence in the room, an inability to move, pressure on the chest, terror, no tracks and no second witness. How large a share of reports take that form has not been counted here, and no published count was found. What has been counted is the underlying experience. Brian Sharpless and Jacques Barber's systematic review of 35 studies (N = 36,533) found lifetime sleep paralysis rates of 7.6 percent in the general population, 28.3 percent in students and 31.9 percent in psychiatric patients, with 34.6 percent among patients with panic disorder. DOCUMENTED

Its content is patterned rather than random. J. Allan Cheyne, Steve Rueffer and Ian Newby-Clark developed and tested a three-factor model of the hallucinations accompanying sleep paralysis. One factor, Intruder, consists of "sensed presence, fear, and auditory and visual hallucinations." A second, Incubus, comprises "pressure on the chest, breathing difficulties, and pain." A third, Unusual Bodily Experiences, covers floating and flying sensations, out-of-body experiences and feelings of bliss. They write that descriptions of these experiences "are remarkably consistent across time and cultures and consistent also with known mechanisms of REM states." DOCUMENTED

David Hufford's 1982 fieldwork on the Newfoundland "Old Hag" made a related argument from folklore, that a stable experience gets named from whatever tradition is locally available, and Owen Davies traced the same experience through early-modern witchcraft accusations. Applying that to bedroom sasquatch reports is this entry's own inference, offered as a hypothesis and nothing more. ALLEGED No published study testing sleep paralysis against bedroom sasquatch accounts was found, and no such study is claimed to exist. Hufford's book was verified through published reviews rather than read.

What the research does and does not license

The asymmetry that actually matters

Everything above establishes that sincere error is possible, and under some conditions likely. It establishes nothing about any particular report. A mechanism capable of producing a false positive does not demonstrate that a given positive is false, and an argument that slides from the first to the second has stopped being an argument about evidence.

The population-level work carries the same limit. Foxon's regression of BFRO sighting counts on black bear populations reported that "every 1,000 bear increase in the bear population is associated with a 4% (95% CI: 1%-7%) increase in sasquatch sightings," and that "across all states and provinces in 2006, after controlling for human population and forest area, there were approximately 5,000 bears per sasquatch sighting." DOCUMENTED The coefficient is the durable part. The 5,000 figure is a fitted average for a single year that moved from roughly 900 to roughly 5,000 across three preprint versions as the model specification changed, and the bear correlation entry works through that history. Nothing in the regression identifies which reports are bears.

The honest summary is uncomfortable in both directions. A person can be truthful, sober, experienced in the woods, unmotivated by attention or money and completely certain, and still be describing something other than what was in front of them. That is a claim about human perception and memory in general, and it covers the person writing the debunk as fully as the person filing the report. Same rewritable memory, same top-down expectation, same habit of experiencing certainty as though it were evidence. Being unimpressed by a sasquatch report runs on the same equipment as being convinced by one, and has earned the same scrutiny.

Sources

Where a full text could not be reached, that is stated. Nothing in this entry is quoted from a document that was not opened.

• Bigfoot Field Researchers Organization, Geographic Database of Bigfoot/Sasquatch Sightings & Reports. Source for the per-state report counts, which change as reports are added. https://www.bfro.net/GDB/
• Bigfoot Field Researchers Organization, "The BFRO's Report Classification System." Source for the Class A and Class B definitions and the sentence on what the system rates. https://www.bfro.net/GDB/classify.asp
• Michael Kalloniatis and Charles Luu, "Light and Dark Adaptation," Webvision: The Organization of the Retina and Visual System. Full text read. Source for the 0.03 cd/m² cone-to-rod crossover, the loss of color under rod vision, the 5 to 10 minute improvement, and the 10⁻⁵ cd/m² absolute threshold at about forty minutes. https://www.webvision.pitt.edu/book/part-viii-gabaergic-mechanisms/light-and-dark-adaptation/
• Dale Purves, George J. Augustine, David Fitzpatrick, et al., eds., "Anatomical Distribution of Rods and Cones," Neuroscience, 2nd ed., NCBI Bookshelf. Full text read. Source for the rod-free 300 µm foveola, the 75 percent acuity reduction at 6° eccentricity, the lower detection threshold outside central vision, and the faint-star sentence. https://www.ncbi.nlm.nih.gov/books/NBK10848/
• Willem A. Wagenaar and Juliette H. Van der Schrier, "Face recognition as a function of distance and illumination: A practical tool for use in the courtroom," Psychology, Crime and Law 2(4), pp. 321-332, 1996. DOI 10.1080/10683169608409787. Deposited abstract read in full via OpenAlex; the article itself is paywalled and was not read. Source for the twelve-second exposure, the six-portrait lineup, the 3 to 40 meter by 0.3 to 3000 lux matrix, and the Rule of Fifteen as the authors state it.
• T. J. Nyman, J. Antfolk, J. M. Lampinen, M. Tuomisto, J. K. Kaakinen, J. Korkman and P. Santtila, "A stab in the dark: The distance threshold of target identification in low light," Cogent Psychology 6(1), 2019. DOI 10.1080/23311908.2019.1632047. Published abstract read verbatim; the publisher's full text returned 403 and was not read. Source for the 178 participants, the three lux conditions, the eight-person target-present lineups, the .53/.41/.11 rates at 20 m, and the authors' stated limitation.
• Jiangang Liu, Jun Li, Lu Feng, Ling Li, Jie Tian and Kang Lee, "Seeing Jesus in toast: Neural and behavioral correlates of face pareidolia," Cortex 53, pp. 60-77, 2014. DOI 10.1016/j.cortex.2014.01.013. Published abstract read verbatim; full text not read. Source for the 34 and 38 percent illusory-report rates, the right fusiform face area finding, and the top-down conclusion.
• Daniel J. Simons and Christopher F. Chabris, "Gorillas in our midst: sustained inattentional blindness for dynamic events," Perception 28(9), pp. 1059-1074, 1999. DOI 10.1068/p281059. Full text read. Source for the 192 retained observers, the 75-second videos, the 5-second unexpected event, the 54 percent overall noticing rate, the 67 versus 42 percent opaque and transparent split, and the 44 versus 65 percent gorilla and umbrella-woman comparison.
• Innocence Project, "DNA Exonerations in the United States." Read. Source for 375 DNA exonerees, 69 percent involving eyewitness misidentification, 5,284 years served, the 14-year average, the 180 exonerees whose real perpetrator was identified, and the 165 actual assailants. https://innocenceproject.org/dna-exonerations-in-the-united-states/
• Innocence Project, "Eyewitness Misidentification." Read. Source for the "more than 60% of our clients" figure and its different denominator. https://innocenceproject.org/eyewitness-misidentification/
• John T. Wixted and Gary L. Wells, "The Relationship Between Eyewitness Confidence and Identification Accuracy: A New Synthesis," Psychological Science in the Public Interest 18(1), pp. 10-65, 2017. DOI 10.1177/1529100616686966. Abstract read in full; article not read.
• John T. Wixted, Laura Mickes and Ronald P. Fisher, "Rethinking the Reliability of Eyewitness Memory," Perspectives on Psychological Science 13(3), pp. 324-335, 2018. DOI 10.1177/1745691617734878. Bibliographic record verified via Crossref. Cited only to identify the paper Wade, Nash and Lindsay were answering.
• Kimberley A. Wade, Robert A. Nash and D. Stephen Lindsay, "Reasons to Doubt the Reliability of Eyewitness Memory: Commentary on Wixted, Mickes, and Fisher (2018)," Perspectives on Psychological Science 13(3), pp. 339-342, 2018. DOI 10.1177/1745691618758261. Abstract read in full; article not read.
• Shari R. Berkowitz, Brandon L. Garrett, Kimberly M. Fenn and Elizabeth F. Loftus, "Convicting with confidence? Why we should not over-rely on eyewitness confidence," Memory 30(1), pp. 10-15. Published online November 2020, in print January 2022. DOI 10.1080/09658211.2020.1849308. Abstract read; article not read.
• Karim Nader, Glenn E. Schafe and Joseph E. LeDoux, "Fear memories require protein synthesis in the amygdala for reconsolidation after retrieval," Nature 406, pp. 722-726, 2000. DOI 10.1038/35021052. Abstract read verbatim via Europe PMC; article paywalled and not read. Source for the reactivation result, the no-reactivation control and the six-hour delay control.
• Elizabeth F. Loftus, "Planting misinformation in the human mind: a 30-year investigation of the malleability of memory," Learning and Memory 12(4), pp. 361-366, 2005. DOI 10.1101/lm.94705. Abstract read verbatim via Europe PMC; full text returned 403 and was not read. Both quoted phrases are from the abstract.
• Alan Scoboria, Kimberley A. Wade, D. Stephen Lindsay, Tanjeem Azad, Deryn Strange, James Ost and Ira E. Hyman, "A mega-analysis of memory reports from eight peer-reviewed false memory implantation studies," Memory 25(2), pp. 146-163, 2017 (online November 2016). DOI 10.1080/09658211.2016.1260747. Abstract read verbatim via Europe PMC; article not read. Source for the 423 participants, 30.4 percent, 23 percent, and the 46.1 percent condition. Note that OpenAlex's reconstruction of this abstract inverts the sense of that last sentence by dropping the word "not"; the publisher-deposited text reads "was not accompanied by a photo."
• Henry L. Roediger III, Michelle L. Meade and Erik T. Bergman, "Social contagion of memory," Psychonomic Bulletin and Review 8(2), pp. 365-371, 2001. DOI 10.3758/BF03196174. Abstract read verbatim via Europe PMC; article not read. Source for the six household scenes, the 15 and 60 second exposures, the confederate procedure, and the schema-consistency and exposure-duration effects.
• Jamie Lewis and Andrew Bartlett, "The Shape of Bigfoot: Transmuting Absences into Credible Knowledge Claims," Cultural Sociology 20(1), pp. 75-95. Published online 5 August 2024, in print March 2026. DOI 10.1177/17499755241264879. Abstract read; the publisher returned 403 and the article was not read. Everything attributed to it here comes from the abstract.
• Brian A. Sharpless and Jacques P. Barber, "Lifetime prevalence rates of sleep paralysis: a systematic review," Sleep Medicine Reviews 15(5), pp. 311-315, 2011. DOI 10.1016/j.smrv.2011.01.007. Abstract read; article not read. Source for the 35 studies, the 36,533 participants, and the 7.6, 28.3, 31.9 and 34.6 percent rates.
• J. Allan Cheyne, Steve D. Rueffer and Ian R. Newby-Clark, "Hypnagogic and hypnopompic hallucinations during sleep paralysis: neurological and cultural construction of the night-mare," Consciousness and Cognition 8(3), pp. 319-337, 1999. DOI 10.1006/ccog.1999.0404. Abstract read verbatim via Europe PMC; article not read. Source for all three factors and the cross-cultural consistency sentence as worded.
• David J. Hufford, The Terror That Comes in the Night: An Experience-Centered Study of Supernatural Assault Traditions, University of Pennsylvania Press, 1982. Not read. Existence, title, subtitle, publisher and year verified through Open Library and through Lorna Amarasingham Rhodes's review in Medical Anthropology Newsletter 15(2), pp. 50-51, 1984, DOI 10.1525/maq.1984.15.2.02a00170.
• Owen Davies, "The Nightmare Experience, Sleep Paralysis, and Witchcraft Accusations," Folklore 114(2), pp. 181-203, 2003. DOI 10.1080/0015587032000104211. Abstract read; article not read.
• Floe Foxon, "If it's there, could it be a bear?", Journal of Zoology 323(1), pp. 1-8, 2024, DOI 10.1111/jzo.13148, and the bioRxiv preprint, DOI 10.1101/2023.01.14.524058. Preprint version 3 full text read. Source for the 4 percent per 1,000 bears estimate with its 1 to 7 percent interval, the roughly 5,000 bears per sighting average, and the statement that the sighting data were "sourced from the Bigfoot Field Researchers Organization (2023) Geographic Database of Bigfoot/Sasquatch Sightings & Reports" for the year 2006. The movement of the ratio across preprint versions is documented in the bear correlation entry.

Cut from an earlier draft of this entry and not replaced: Loftus and Palmer (1974), Journal of Verbal Learning and Verbal Behavior 13, pp. 585-589. The bibliographic record resolves, but no reachable copy of the article was found, so nothing is attributed to it here.