De wittenskip achter wêrom desinformaasje flugger reizget as de wierheid
April 4, 2025 · Frisian News
Research shows false claims spread three times faster on social media than accurate information, driven by human psychology rather than algorithmic bias alone. Scientists now understand the mechanics behind this pattern, though fixing it remains difficult.
FryskIn leugen reizget fan New York nei Tokio wylst de wierheid noch har skuon fêstmakket. Dit âlde sprekwurd hat no sifers achter him. Ûndersikers oan MIT en Stanford bestudearren 126.000 Twitter-threads tusken 2006 en 2017 en fûnen dat ûnwiere bewearingen 1.500 minsken seis kear flugger berikten as wiere. De fersprieding sels seach der ek oars út. Wierheid groeide as in stadich fjoer, wûn oanhingers fia bestindige dielen. Leugens eksplodearren ynienen nei bûten, griepen oandacht en riepen gau lilkens op.
Minsken, net algoritmes, driuwe dizze motor oan. Doe't sosjale mediabedriuwen yn resinte jierren stopten mei it pushen fan ynhâld puer op basis fan belutsenheid, fersprieden ûnwiere bewearingen him noch altyd flugger. Us harsens evoluearren om nijichheid en bedriging op te merken. Ûnwierheid biedt faak nijichheid omdat it tsjinsprekt wat wy al leauwe. In bewearing dy't dyn wrâldbyld tsjinsprekt, feroarsaket emoasje. Do wolst it diele om oaren te warskôgjen, of om te argumentearjen. In saaie, fertrouwde wierheid oer ynflaasje of hannelepolityk stekt dat fjoer net oan. Do scrollest der lâns.
De psychologyske trúk wurket yn alle rjochtingen. Rjochtsige netwurken ferspriede falske bewearingen oer ferkiezingsfraude. Linksige netwurken ferspriede ûnwiere bewearingen oer bedriuwskea. Beide partijen fiele dat se wierheid tsjin leugens ferdigenje. Beide kanten diele flugger as se moreel ûntstemd fiele. De harsens meitsje gjin ûnderskied tusken politike stammen as it om dizze reaksje giet. In âld ûndersyk toande oan dat minsken ûnwiere bewearingen better ûnthâlde as se se wjerlizze. Do tinkst dat do de feiten rjochtsetst. Ynstee dêrfan ferankerje do de leugen yn it geheugen.
Techbedriuwen besochten dit op te lossen troch ûnwiere ynhâld oan te jaan, kontekstlabels ta te foegjen of akkounts te ferwiderjen. Dizze stappen fertrage guon fersprieding, mar feiten bleauwen koppich. Do kinst immen net mei logika út in posysje helje dy't se net logysk yn gien binne. In label dat seit dat eat ûnwier is, ferstevighet faak it leauwen yn de leugen. Minsken sjogge it label as bewiis fan ferberging. Se fertroue de ûnwiere bewearing noch mear. Facebook ûntdekte dit doe't it warskôgingen oan misleidende berjochten oer faksins tafoege. De warskôgingen wurken allinnich as minsken al oan de ûnwiere bewearing twiifelen.
Wierheid hat bûnsgenoaten nedich, net allinnich bettere algoritmes. Echte mienskippen mei in mienskiplik belang by krektens ferslaan isolearre feitenkontrôleurs. As in lokale sjoernalist eltse wike stedsfergeadrings ferslacht, leare bewenners him te fertrouwen. As in oerheidsynstânsje rapporten útjout dy't nimmen lêst, ûntstiet der gjin fertrouwen. Lytse groepen dy't oan echte problemen yn har eigen plakken wurkje, sprekke lûder as ekspertpanels dy't út haadsteden roppe. It antwurd sit net yn technology mar yn it drege wurk fan it opbouwen fan lokale leauwichheid, ien ferhaal tagelyk.
A lie travels from New York to Tokyo while truth still buckles its shoes. This old saying now has numbers behind it. Researchers at MIT and Stanford studied 126,000 Twitter threads between 2006 and 2017 and found that false claims reached 1,500 people six times faster than true ones. The spread itself looked different too. Truth grew like a slow burn, gaining followers through steady shares. Lies exploded outward in sudden spikes, grabbing attention and triggering anger in quick succession.
People, not algorithms, drive this engine. When social media companies stopped pushing content based purely on engagement in recent years, false claims still spread faster. Our brains evolved to notice novelty and threat. Falsehoods often offer novelty because they contradict what we already believe. A claim that contradicts your worldview sparks emotion. You want to share it to warn others, or to argue. A boring, familiar truth about inflation or trade policy does not light that fire. You scroll past it.
The psychological trick works in all directions. Right-leaning networks spread false claims about election fraud. Left-leaning networks spread false claims about corporate harm. Both sides feel they are defending truth against lies. Both sides share faster when they feel moral outrage. The brain does not distinguish between political tribes when it comes to this reaction. An old study showed that people remember false claims better when they refute them. You think you are correcting the record. Instead you cement the lie in memory.
Tech companies tried fixing this by flagging false content, adding context labels, or removing accounts. These steps slowed some spread, but facts remained stubborn. You cannot logic someone out of a position they did not logic themselves into. A label saying "this is false" on a lie often hardens belief in the lie. People see the label as proof of a cover-up. They trust the false claim even more. Facebook found this when it added warnings to misleading posts about vaccines. The warnings worked only when people already doubted the false claim.
Truth needs allies, not just better algorithms. Real communities with shared stake in accuracy beat isolated fact-checkers. When a local journalist covers town hall meetings every week, residents learn to trust them. When a government agency churns out reports no one reads, no trust forms. Small groups working on real problems in their own places speak louder than expert panels shouting from capitals. The answer sits not in technology but in the hard work of building local credibility, one story at a time.
Published April 4, 2025 · Frisian News · Ljouwert, Fryslân