Fast impressions
Judgments of a face made in 100 ms closely matched judgments made with no time limit.
Willis & Todorov (2006), Psychological Science 1Science
Every claim on the home page traces to a study here, stated at its published size and with its limits.
Findings from peer-reviewed studies, stated at their published size, with their limits.
Judgments of a face made in 100 ms closely matched judgments made with no time limit.
Willis & Todorov (2006), Psychological Science 1Raters agree on who is attractive, within and across cultures.
Langlois et al. (2000), Psychological Bulletin 2Attractive people are judged and treated more positively, even by people who know them.
Langlois et al. (2000), Psychological Bulletin 2Fast impressions are consistent, not necessarily correct. Extra viewing time mostly added confidence.
Willis & Todorov (2006), Psychological Science 1Across experimental studies, attractive candidates were favored. A moderate effect: d = 0.37.
Hosoda et al. (2003), Personnel Psychology 3The same meta-analysis found that the bias decreased over time.
Hosoda et al. (2003), Personnel Psychology 3In surveys from the 1970s and 1980s, plainness carried a 5 to 10% pay penalty. An association, not the effect of changing your looks.
Hamermesh & Biddle (1994), American Economic Review 4In a lab labor market, part of the beauty premium ran through the workers' own confidence.
Mobius & Rosenblat (2006), American Economic Review 5In community speed dating, both sexes chose mainly on physical attractiveness.
Asendorpf et al. (2011), European Journal of Personality 6Only 4% of pairs formed a relationship within the following year.
Asendorpf et al. (2011), European Journal of Personality 6In a field experiment with 3,000 dating-app users, expansive postures in photos were chosen more often.
Vacharkulksemsuk et al. (2016), PNAS 7Among 8,520 women, beards raised ratings for long-term relationships, not short-term ones.
Dixson et al. (2016), Journal of Evolutionary Biology 8In men's skin, more even pigment and redness read as younger, healthier and more attractive.
Fink et al. (2012), JEADV 9A 4.5-year randomized trial found 24% less skin aging with daily use than with occasional use.
Hughes et al. (2013), Annals of Internal Medicine 10A small study linked eating more of them to a healthier-looking skin tone within about six weeks.
Whitehead et al. (2012), PLoS ONE 11Asked to make faces look healthiest, people raised the rosy, oxygenated-blood color above natural levels.
Stephen et al. (2009), PLoS ONE 12After 31 hours awake, raters saw heavier eyelids, darker circles and paler skin.
Sundelin et al. (2013), Sleep 13Rated attractiveness fell by about 2 points out of 100 after a night without sleep.
Axelsson et al. (2010), BMJ 14After two nights of restricted sleep, raters were less willing to socialize with the people photographed.
Sundelin et al. (2017), Royal Society Open Science 15The largest study, with 181 people, found no objective facial change after one lost night.
Holding et al. (2019), Journal of Sleep Research 16At 12 inches, the base of the nose appears about 30% wider than at 5 feet. A geometric model.
Ward et al. (2018), JAMA Facial Plastic Surgery 17Photos of the same person varied in attractiveness more than photos of different people did.
Jenkins et al. (2011), Cognition 18Which photo of a person was shown biased impressions, even at a 40 ms glance.
Todorov & Porter (2014), Psychological Science 19Expression was the largest source of variation between photos of the same person.
Sutherland et al. (2017), British Journal of Psychology 20The evidence
Ten of the papers behind Klayr, each with its design, size and main limitation.
Hughes et al., 2013. Daily use, 24% less aging over 4.5 years. Limit: a high-UV Australian setting; prevention, not reversal.10
Holding et al., 2019. No objective change after one lost night. Limit: one night only.16
Axelsson et al., 2010. About 2 points out of 100 after 31 hours awake. Limit: small and extreme.14
Jenkins et al., 2011. Photo-to-photo variation exceeded person-to-person. Limit: uncontrolled photos, by design.18
Ward et al., 2018. About 30% wider nasal base at 12 inches. Limit: modeled, not measured in people.17
Fink et al., 2012. Even pigment and redness read as healthier. Limit: Caucasian sample, cheek crops.9
Whitehead et al., 2012. Skin color shifted within six weeks. Limit: small, observational.11
Hosoda et al., 2003. A moderate bias, shrinking over time. Limit: simulated decisions.3
Rhodes, 2006. Symmetry plays a modest part and is set by bone. Limit: mostly Western samples.21
Koran et al., 2008. 2.4% of adults screened positive. Limit: lay-interviewer screen.25
Method
Four commitments that decide whether a number is shown to you at all.
Fixed distance, even light and a level head, following clinical imaging guidance.24
Before launch, a repeat-scan study will measure how much each reading moves between photos taken minutes apart.
Only what habits can change. Nothing fixed by bone is scored.
Commercial face analysis has erred far more on darker skin. Every measure will be audited across skin tones before launch.26
Each one responds to habits. Symmetry and bone proportions do not, so Klayr does not measure them.21
Re-measured weeklyFrom home
Statistics are averages from the cited studies. Individual results vary.
You're on the list.
What would you most like to work on? Optional.