PNAS：美國公眾對(duì)科學(xué)的信任趨勢(shì) Trends in US public confidence in science and opportunities for progress

Downloadedfromby

26

onApril1,2024fromIPaddress

26

.

ATTHE

NATIONAL

ACADEMIES

OPENACCESS

TrendsinUSpublicconfidenceinscienceandopportunitiesforprogress

ArthurLupiaa,1,2,DavidB.Allisonb,3,KathleenHallJamiesonc,2,JenniferHeimbergd,4,MagdalenaSkippere,2,andSusanM.Wolff,2

Inrecentyears,manyquestionshavebeenraisedaboutwhetherpublicconfidenceinscienceischanging.Toclarifyrecenttrendsinthepublic’sconfidenceandfactorsthatareassociatedwiththesefeelings,aneffortinitiatedbytheNationalAcademies’StrategicCouncilforResearchExcellence,Integrity,andTrust(theStrategicCouncil)analyzedfindingsfrommultiplesurveyresearchorganizations.TheStrategicCouncil’seffort,whichbeganin2022,foundthatU.S.publicconfidenceinscience,thescientificcommunity,andleadersofscientificcommunitiesishighrelativetoothercivic,cultural,andgovernmentalinstitutionsforwhichresearchersregularlycollectsuchdata.However,confidenceintheseinstitutionshasfallenduringtheprevious5years.Science’sdecline,whilereal,issimilartoorlessthanthatintheothergroups.Arecentstudygoesintogreaterdetailbyexploringpublicviewsofscience.Fromthesedata,weobservethatmanyofthesurveyedU.S.publicquestiontheextenttowhichscientistssharetheirvaluesorovercomepersonalbiaseswhenpresentingconclusions.Atthesametime,largemajoritiesagreeoncertaintypesofactionsthattheywantscientiststotake.Forexample,84%respondthatitis“somewhatimportant”or“veryimportant”forscientiststodisclosetheirfunders.Ninety-twopercent(92%)offerthesameresponsestoscientists“beingopentochangingtheirmindsbasedonnewevidence.”Collectively,thesedataclarifyhowtheU.S.publicviewsscienceandscientists.Theyalsosuggestactionsthatcanaffectpublicconfidenceinscienceandscientistsintheyearstocome.

Science’scapacitytoproducediscoveriesthathelppeoplebetterunderstandcriticalaspectsoftheiruniverse,theenvi-ronmentsinwhichtheylive,andoneanotherisundisputed.Scienceisatthecoreoftransformativetechnologiesandinnovativenewmaterialsandpracticesthatimprovehealth,increaseeconomicopportunity,andenhancethequalityoflifeforpeoplearoundtheworld.Somescientificpursuitsalsoprovokesocietalcontroversy.Topicssuchasclimatechangeandvaccinesafetynotonlysparkdebatebutalsoleadsomepeopletoquestiontheintegrityofthescienceitself.

Phenomenasuchastheseyieldheadlinessuchas“Canthepublic’strustinscience—andscientists—berestored?”(1).Indeed,inrecentyears,media,individuals,andscientificorganizationshaveexpressedarangeofopinionsaboutwhetherpublicconfidenceinscienceisdecliningandhaveofferedconjectureandevidenceaboutfactorsunderlyingpublicconfidenceinscience(2,3)andaffectingsupportforfundingit(4).TheNationalAcademies’StrategicCouncilforResearchExcellence,Integrity,andTrust(theStrategicCouncil)initiatedthisstudytoexaminebothpublicconfi-denceandthefactorsaffectingit.

TheStrategicCouncilwasformedin2021(5).Itsinitialactiv-itieshaveincludedexaminingways:tomakepotentialconflictsofinteresteasiertoidentifyanddisclose,toimproveincen-tivesassociatedwithcorrectingthescientificrecordwhenmistakesarefound(thatis,improvingtheretractionprocess),toevaluateframeworkstoincreaseincentivesforscientificintegrity,toexplorehowresearchlaboratoriesandlargersci-entificgroupscanmoreeffectivelyintegratescientificintegrityandresearchethicsintotheirscientificpractice,andtoassesspublicconfidenceinscience(thepurposeofthisarticle).Ingeneral,theStrategicCouncilisseekingwaystoworkwithawiderangeofpartnerstosupportscientificexcellencewhilepromotingpracticesthatstrengthenintegrityandreducebureaucraticburdensforresearchersandinstitutions.

Inthisarticle,wepresentevidenceofchangesinpublicconfidenceinsciencefromtwosources.First,wesynthesizetrenddatafromhigh-qualitysurveyresearchorganizations.Byhighquality,wemeanresearchfirmsthatpubliclycommittoasetofdatacollectionandinterpretationpracticesthatincreasethelikelihoodofaccurateinterpretations.Second,wedescribefindingsfromanewlydevelopedapproachtoanalyzingpublicconfidenceinsciencebyaskingdetailedquestionsnotjustaboutscienceingeneralbutalsoaboutperceptionsbothofscientists’adherencetothescientificnormstheyespouseandoftheincentivesthatmotivateindi-vidualscientistsandorganizations.

Fromthesedata,weconcludethat

1.Confidenceinscienceishighrelativetonearlyallothercivic,cultural,andgovernmentalinstitutionsforwhichdataarecollected,aconclusionconsistentwithlong-termtrends.

Authoraffiliations:aOfficeoftheVicePresidentforResearchandDepartmentofPoliticalScience,UniversityofMichigan,AnnArbor,MI48109;bDean,SchoolofPublicHealth,IndianaUniversity,Bloomington,IN47405;cAnnenbergPublicPolicyCenter,UniversityofPennsylvania,Philadelphia,PA19104;dPolicyandGlobalAffairsDivision,TheNationalAcademiesofSciences,Engineering,andMedicine,Washington,DC20001;eNature,LondonN19XW,UnitedKingdom;andfUniversityofMinnesotaLawSchoolandMedicalSchool,Minneapolis,MN55455

Authorcontributions:A.L.,D.B.A.,andK.H.J.designedresearch;A.L.andK.H.J.performedresearch;A.L.performeddatachecks,compiledreferences,andledthewritingprocess;D.B.A.contributedideasonstatisticalinferenceandrevieweddrafts;K.H.J.revieweddraftsandcompiledreferences;J.H.revieweddraftsanddatachecks;M.S.andS.M.W.contributedideasandrevieweddrafts;andA.L.,K.H.J.,J.H.,M.S.,andS.M.W.wrotethepaper.

Theauthorsdeclarenocompetinginterest.

Copyright?2024theAuthor(s).PublishedbyPNAS.Thisopenaccessarticleisdistributedunder

CreativeCommonsAttributionLicense4.0(CCBY)

.

1Towhomcorrespondencemaybeaddressed.Email:

lupia@

.

2MemberoftheStrategicCouncilforResearchExcellence,Integrity,andTrust.3Co-chairoftheStrategicCouncilforResearchExcellence,Integrity,andTrust.4DirectoroftheStrategicCouncilforResearchExcellence,Integrity,andTrust.

Thisarticlecontainssupportinginformationonlineat

/lookup/

suppl/doi:10.1073/pnas.2319488121/-/DCSupplemental

.

PublishedMarch4,2024.

PNAS2024Vol.121No.11e2319488121

/10.1073/pnas.2319488121

1of9

Downloadedfromby

26

onApril1,2024fromIPaddress

26

.

2.Confidencehasdeclined…butthedeclineisnotscience-specific.Instead,science’sdeclineissimilartoorlesspro-nouncedthanconfidenceinmanyinstitutions.

3.AsofFebruary2023,thepublichashighlevelsofconfidenceinscientists’competence,trustworthiness,andhonesty.Forexample,whenaskedthequestion“Howconfidentareyouthatscientistsprovidethepublicwithtrustworthyinfor-mationaboutthescienceintheirareaofinquiry?”84%ofrespondentsreportthattheyareveryconfidentorsome-whatconfident(see

SIAppendix

).

4.However,manyU.S.adultsquestionwhetherscientistssharetheirvaluesandwhethertheycanovercometheirbiases.Forexample,whenasked,“Whenastudyrunscountertotheinterestsoftheorganizationrunningthestudy,whichismorelikelytohappen?”andgivenachoicebetweentheresponse“Scientistswillpublishthefinding”and“Scientistswillnotpublishthefinding”70%ofthesamplechoosesthelatter(see

SIAppendix

).

5.Thepublichasconsistentbeliefsabouthowscientistsshouldactandbeliefsthatsupporttheirconfidenceinsciencedespitetheirconcernsaboutscientists’possiblebiasesanddistortiveincentives.Eighty-fourpercent(84%)ofU.S.adultsrespondedthatitis“somewhatimportant”or“veryimportant”forscientiststodisclosetheirfunders.Ninety-twopercent(92%)offeredthesameresponsestoaquestionabouttheimportanceofscientists“beingopentochangingtheirmindsbasedonnewevidence”(see

SIAppendix

).

Collectively,thesedataclarifyhowmembersofthepublicviewscienceandscientistsandrevealdetailsaboutwhattypesofactionscouldaffecttheirconfidenceinscienceandscientistsgoingforward.

Weproceedasfollows.First,wedescribeourdataselec-tioncriteria.ManysurveyscontainquestionsaboutpublicconfidenceinsciencebutnotallhewtowidelyrecognizedbestpracticesformakingrepresentativeclaimsabouttheU.S.population.Next,weofferanoverviewoftrendsinpub-licconfidenceinsciencedatingbacktwentyyears.Tohigh-lighttheimportanceofscienceconfidencetrends,weshowresultsfromastudythatshowedhowvariationsinconfi-dencecorrespondedtoU.S.adults’decisionstotakeoneoftheCOVID-19vaccines.Then,weuseresultsfromanotherstudythatprovidesmoredetaileddatatorevealahigherlevelofnuanceinpublicviewsofscienceandscientists.Inthefinalsection,weuseinsightsfromthesedatatodescribestepsthatresearchersandresearchorganizationswhoareaddressingU.S.audiencescantaketoincreaseconfidenceintheirpracticesandfindingswhensuchconfidenceiswar-ranted.Takingthesestepsmayincreasepublicacceptanceoftheirresearch.

DataSelectionandAttributes

Surveyresearchprovidestheempiricalcorpusfromwhichmostofthisreviewdraws.Surveysaskingquestionsrelatingto“confidence”insciencevaryinquality.Werestrictouratten-tiontoresearchfromsurveyorganizationsthatcollectdatatofacilitatein-depthresearchandthatfollowbestpracticescon-cerningscientificsamplingprocedures.Fortime-trendpres-entationsthatcharacterizelevelsoftrustinscienceinthe

UnitedStates,weusedatafromnationallyrepresentativesurveysofthepopulationfromproducersthatadopttheAmericanAssociationforPublicOpinionResearch(AAPOR)CodeofProfessionalEthicsandPractices,whoaresignatoriesofAAPOR’sTransparencyInitiativeorwhohavemadeequivalentmethodologicalcommitments,atthetimeofdatacollection(6).Inotherpartsofthisarticle,weciteresearchthatdocu-mentscorrelatesofscientificattitudes.Thesestudiesexaminespecificrelationshipsanddonotmakerepresentativeclaimsaboutthenation.Thesesupplementarystudiesalsousewell-documentedandpubliclyaccessiblemethodologies.

Wheninterpretingsurveyresearchfindings,thewordingofquestionsmatters.Foreachfindingbelow,wepresenttheexactquestionwordingthatelicitedtheresponse.Researcherssometimeswordquestionsindifferentways.Forexample,someaskaboutconfidencein“science”whileothersfocusonconfidencein“thescientificcommunity”andstillotherscon-centrateon“trust.”Thesevariationsareanormalpartofsurveyresearch,reflectingthefactthatresearchershavediverseinter-estsandmeasuredifferentphenomena.Inmanymediaandotherpublicconversationsaboutconfidenceinscience,thereisafocuson“trustinscience”andconcernsthat“trustisfalling.”Whilerecognizingthattrustappearsmorefrequentlyincom-monparlance,thetermconfidencemoreaccuratelyreflectsthequestionsthatprominentsurveyresearchorganizationsask.Wewillsaymoreaboutwhatfindingsaboutconfidenceimplyabouttrustinscienceinthearticle’sfinalsection.

ScienceConfidenceTimeTrends

TomeasureconfidenceinU.S.civicinstitutions,thePewResearchCenterasks,“Howmuchconfidence,ifany,doyouhaveineachofthefollowingtoactinthebestinterestsofthepublic?”Foreachcivicinstitution,surveyparticipantscanselectoneofthefollowingresponses,“agreatdealofconfi-dence,”“afairamountofconfidence,”“nottoomuchconfi-dence,”“noconfidenceatall,”ortheycanchoosenottoanswer.Pewasksthisquestionaboutmanygroupsincludingscientists,medicalscientists(Pewisoneofonlyafewsurveyresearchorganizationsthatasksthesequestionsseparately),themilitary,policeofficers,religiousleaders,journalists,electedofficials,andmore.InFig.1,wedrawfromPew’srecordedconfidencechangesfrom2016to2023(2).

Thefigure’sy-axisreferstothepercentageofrespondentswhoanswerthequestioninaspecificway.Thedarkbluepartsofeachhorizontalbarrefertothepercentagerespond- ing,“agreatdealofconfidence.”Thelightbluepartrepresents thepercentageresponding,“afairamountofconfidence.”Thegreenpartreferstothecombinedpercentagewhoreport“nottoomuch”or“noconfidenceatall”inthenamedcivic institution.

Fig.1showshighlevelsofconfidencein“scientists”and“medicalscientists”relativetoothergroups.Theselevelsarenearlyidenticaltoconfidencein“themilitary”andhigherthan thatforpoliceofficers,religiousleaders,journalists,business leaders,andelectedofficials.ThesedataalsoshowthatU.S.publicconfidenceinscientistsandmedicalscientistsdropped from2020to2023.Thisdecline,however,isnotunique.Allmeasuredinstitutionsexperienceddeclinesoverthesameperiodwithmostsimilarinsizetothedeclineinscience.

2of9

/10.1073/pnas.2319488121

Downloadedfromby

26

onApril1,2024fromIPaddress

26

.

84

74

73

131114

15

17131215

222022

2321222327

JunFebDecJanAprNovDecsepoct

'16'18'18'19'20'20'21'22'23

252326

JunFebDecJanAprNovDecsepoct

'16'18'18'19'20'20'21'22'23

JunJanAprNovDecsepoct

'16'19'20'20'21'22'23

policeofficers

Religiousleaders

NET7878

7775

74

70

6163

59

555353

222126

313031

353134

AprNovDecsepoct

'20'20'21'22'23

Dec

DecJanAprNovDecsepoct

'18'19'20'20'21'22'23

Electedofficials

Businessleaders

55

41444346484640

44

42

40

37353737

35

28

2725

24

63

72

63

64

62

75

71

75

64

58

54

44

52

55

58

201917181617

publicschoolprincipals

8083

5349

57

465038433740454746

JunFebDecJanAprNovDecsepoct

'16'18'18'19'20'20'21'22'23

NET

Journalists

4845

646865

20221725

24

39

MajoritiesofAmericanssaytheyhaveatleastafairamountofconfidenceinscientists,butratingshavefallensinceearlyinthecoronavirusoutbreak

%ofU.S.adultswhohave__ofconfidenceinthefollowinggroupstoactinthebestinterestsofthepublic

●AgreatdealAfairamount●Nottoomuch/Noconfidenceatall

Medicalscientists

scientists

Themilitary

87

85

87

84

79

78

77

77

77

76

77

74

NET798083828383

76

sep

'22

JunFebDecJanAprNovDecsepoct

'16'18'18'19'20'20'21'22'23

oct

'23

Apr

'20

Dec

'21

Nov

'20

JunFebDecJanAprNovDecsepoct

'16'18'18'19'20'20'21'22'23

Note:Respondentswhodidnotgiveananswerarenotshown.

Source:SurveyofU.S.adultsconductedSept.25-Oct.1,2023.

“Americans’TrustinScientists,PositiveViewsofScienceContinuetoDecline”

PEWRESEARCHCENTER

Fig.1.RecentU.S.trendsinpublicconfidenceinscientists,medicalscientists,andotherinstitutions(2).

InitsGeneralSocialSurvey,theNationalOpinionResearchCenterattheUniversityofChicago(NORC)studiesasimilarphenomenon.WherePewaskedaboutscientificinstitutions,

NORCasksabouttheindividualsrunningscientificinstitu-tions.Specifically,“Iamgoingtonamesomeinstitutionsinthiscountry.Asfarasthepeoplerunningtheseinstitutions

PNAS2024Vol.121No.11e2319488121

/10.1073/pnas.2319488121

3of9

Downloadedfromby

26

onApril1,2024fromIPaddress

26

.

Fig.2.OverallpublicconfidenceinthescientificcommunityintheUnitedStates(7).Imagecredit:CieraHammond.

areconcerned,wouldyousayyouhaveagreatdealofcon-fidence,onlysomeconfidence,orhardlyanyconfidenceatallinthem?”WherePewasksaboutconfidencein“scientists,”NORCasksaboutconfidencein“thescientificcommunity.”Sinceourgoalistoshowapples-to-applescomparisons,weneedeachorganizationtoaskthequestioninaninternallyconsistentwayovertime.ThePewandNORC,foreachoftheirsurveys,satisfythiscondition.

Fig.2revealsthetrendinU.S.publicconfidencechangesfrom2000to2022(7).Init,thedottedbluelinereferstothepercentageresponding,“agreatdealofconfidence,”thesolidredlinerepresentsthepercentageresponding,“onlysomeconfidence”andthesolidorangelinereferstothepercent-agewhoreport“hardlyanyconfidenceatall”inthescientificcommunity.Fig.2showsthatconfidenceinthepeoplerun-ningthescientificcommunityhasbeenhighoverthelasttwodecades.Evenwithasharpdeclinein2022,over85%ofU.S.adultsreporthaving“agreatdeal”or“someconfidence”inthescientificcommunityeveryyearofthesurvey.Krauseetal.(8)andBradyandKent(9)findsimilartrendsoverdif-ferentperiodsusingcomparablesourcesofdata.

Collectively,thesurveydataofferevidencethatwithintheUnitedStates:

?Confidenceinscientistsandtheleadersofthescientificcommunityishighrelativetoothergroups.

?Thisconfidencehasdeclinedinrecentyears,butthescience-relateddeclineiscomparableto,orlesspro-nouncedthan,declinesinconfidenceinothergroups.

AssociationwithCOVID-19VaccinationStatus

Confidenceinscienceprovidesareasonforpeopletopayattentiontoscientificfindings.Thisconfidencebecomesmoreimportantwhenpeopleareaskedtoweighscientificevidence.Acaseinpointisthecorrespondencebetweenpeople’sviewsofscienceandtheirwillingnesstotakealife-savingvaccine.Surveyresearchersfromseveralorgan-izationsexaminedthisrelationship.Eachfoundimportant

relationshipswithintheU.S.adultpopulationbetweenascience-basedformoftrustandwillingnesstotakeaCOVID-19vaccine.

Forexample,theJamiesonetal.(10)surveyempaneledcitizensofFlorida,Michigan,Ohio,Pennsylvania,andWisconsinseventimesbetweenApril2020andMarch2021.Abouttrust,theyaskedthreequestions:

?“Howmuch,ifatall,doyoutrusttheleadersofinstitutionssuchastheU.S.CentersforDiseaseControlandPrevention(CDC)andtheNationalInstitutesofHealth(NIH)toactinthebestinterestofpeoplelikeyou?”

?“Howmuch,ifatall,doyoutrustwhatDr.AnthonyFaucioftheNationalInstitutesofHealth(NIH)tellsyouaboutthecoronaviruspandemic?”

?“Howmuch,ifatall,doyoutrustwhattheU.S.CentersforDiseaseControlandPrevention(CDC)tellsyouaboutthecoronaviruspandemic?”

Foreachquestion,responseoptionswere“agreatdeal,”“alot,”“amoderateamount,”“alittle,”and“notatall.”They thenaggregatedresponsestothesequestionstoforma“trustinhealthauthorities”measure.Aboutwillingnesstovaccinate,theyasked,“Ifano-costvaccinethatprotectspeo-plefromthecoronavirus,alsoknownasCOVID-19,becomesavailableandisapprovedbytheFoodandDrugAdmin- istration,alsoknownastheFDA,howlikely,ifatall,wouldyoubetogetvaccinated?”Responseoptionswere“notatall likelytogetvaccinated,”“nottoolikely,”“somewhatlikely,”and“verylikelytogetvaccinated.”

Fig.3depictstherelationshipbetweenthesevariablesfromJuly2020toFebruary2021.Thefigurereflectstheirmainfinding,whichisthattheU.S.adultpopulation’strustinhealthauthoritieswasasignificantpredictoroftheirreportedintentiontovaccinate.Allingtonetal.(11)reportsimilarcorrespondencesintheUnitedKingdom.

Inaddition,Jamiesonetal.(10)andtheCOVIDStatesProject(12)alsofoundthathigherlevelsoftrustinhealthauthoritieswerestronglyandnegativelycorrelatedwithsubsequent

4of9

/10.1073/pnas.2319488121

Downloadedfromby

26

onApril1,2024fromIPaddress

26

.

VaccineIntentions

0.00.20.40.60.81.0

TrumpRevealsCOVID

ElectionDay

PfizerApproval

BidenTakesOffice

VaccineIntentionsOverTimeByTrustinHealthAuthorities

Notatall(10%)

Alittle(12%)

Amoderateamount(30%)

Alot(21%)

Agreatdeal(26%)

Jul'20Aug'20Sep'20Oct'20Nov'20Dec'20Jan'21Feb'21Mar'21

Date

Fig.3.CorrespondencebetweentheU.S.adultpopulation’strustinhealthauthoritiesandintentiontovaccinate.SurveydatawerecollectedindifferentwavesfromJuly,August/September,September/October,October/November,November,andDecemberof2020,andJanuary/Februaryof2021.Theleft-handgrayregionistheSeptember/October2020waveandtheright-handgrayregionistheJanuary/February2021wave(10).

acceptanceofCOVID-relatedmisinformation.Inotherwords,peopleintheUnitedStateswithlowerlevelsoftrustweremorelikelytoacceptmisinformation,which,inturn,wasasso-ciatedwithalowerreportedwillingnesstovaccinate.

GiventhedemonstratedeffectivenessofCOVID-19vacci-nationatsavinglives,mitigatinghospitalization,andreducingmanyrelatedpublichealthrisks,thesedataofferevidenceofhowlevelsinpublicconfidenceinsciencecorrespondtorealquality-of-lifeoutcomes.

Collectively,theresearchdescribedinthissectionoffersevidencethatwithintheUnitedStates:

?LowerlevelsoftrustinthepublichealthinstitutionsandspokespersonsthatcommunicatehealthsciencewereassociatedwithacceptanceofmisinformationaboutCOVID-19andCOVID-19vaccines.

?AcceptanceofmisinformationaboutCOVID-19vaccineswasassociatedwithdecisionsnottotakealife-savingvaccine.

?HigherlevelsoftrustinthepublichealthinstitutionsandspokespersonsthatcommunicatehealthsciencewereassociatedwithtakingaCOVID-19vaccine.

ACloserLookatUnderlyingFactors

Inmostcases,whenpeopleareaskedtoconsidercondition-ingtheiractionsorbehaviorsonascientificfinding,theyarenotsimultaneouslyofferedtheunderlyingdata,code,orsupplementarymaterialsthatresearcherswouldusetoeval-uateascientificclaim.Inmanycases,thepublicisaskedtobasetheiracceptanceofascientificclaimontrust—trustinmethods,processes,people,orinstitutions.Wintterlinetal.(13)acknowledgethechallengefacingthosewhoarethinking

aboutwhethertoconditiontheirdecisionsonascientificfindingwhentheywrite(pp.1–2):

Scientists(andscienceasawhole)provideevidenceandadviceforsocietalproblemsolvingandcollectivedecision-making.Forthisadvicetobeheard,thepub-licmustbewillingtotrustscience,where“trust”meansthatonecanconfidentlyexpectsciencetoprovidereliableknowledgeandevidence…Becauseoftheirboundedunderstandingofscience,citizensinevitablymusttrustinscience(orscientistsasrep-resentativesofthatsystem),eventhoughthismightberisky…

TogainabetterunderstandingofwhyU.S.adultsvaryintheirwillingnesstotakethiskindofrisk,theAnnenbergPublicPolicyCentersurveyedanempanelednationallyrep-resentativesampleofU.S.adults(see

SIAppendix

formeth-odologicaldetails).Thisstudy,calledtheAnnenbergScienceKnowledgesurvey,orASK,offersawayforresearcherstodistinguisharangeofpublicviewsaboutscienceandscien-tists.TheASKsurveyposedquestionsnotjustaboutscienceandscientistsingeneral,butalsoexaminedwhetherthepublicviewedscientistsatuniversitiesdifferentlythanscien-tistswhoworkforthefederalgovernmentorinindustry.WhiletheASKsurveyincludedanexpansivesetofvariables,wefocushereonthosethatpertaintoconfidenceinscienceorscientistsingeneral.DatareportedbelowarefromanASKsurveyconductedbetweenFebruary22andFebruary28,2023,onarepresentativesampleof1,638empaneledU.S.adults.

Thesurveyinitiallyasked,“Ingeneral,howconfidentareyouthatscientistsprovidethepublicwithtrustworthyinformationaboutthescienceintheirareaofinquiry?”Theresponse

PNAS2024Vol.121No.11e2319488121

/10.1073/pnas.2319488121

5of9

Downloadedfromby

26

onApril1,2024fromIPaddress

26

.

Table1.PublicperceptionofscientistsintheUnitedStates

%strongly%somewhat%neitheragree%somewhat%stronglyNet

ScientistsingeneralagreeagreeNetagreenordisagreedisagreedisagreedisagree

Arecompetent

35

46

81

14

3

2

5

Aretrustworthy

25

45

70

21

7

2

9

Arehonest

22

47

68

21

8

3

10

Areethical

23

43

65

25

8

2

10

Careaboutthewell-

beingofothers

24

44

68

24

6

2

8

Sharemyvalues

11

31

42

45

10

3

13

Feelsuperiortoothers

9

27

36

43

15

6

21

Arelikeable

9

33

42

51

1

5

7

Insomecases,cellsdonotsumto100%duetorounding.Boldvaluesindicateasummationofothercolumns(see

SIAppendix

formethodologicaldetails).

optionswere“veryconfident,”“somewhatconfident,”“nottooconfident,”“notconfidentatall,”and“don’tknow.”Thirty-eightpercent(38%)reportedthattheywere“veryconfident.”Forty-sixpercent(46%)reportedbeing“somewhatconfident.”Fourteenpercent(14%)reportedbeing“nottooconfident”and2%reportedbeing“notconfidentatall.”O(jiān)verall,84%ofrespondentsreportedbeing“somewhat”or“very”confident.

Table1showsresponsestoamoredetailedsetofques-tionsonhowthepublicperceivesscientists.Thissetofques-tionsbeganwiththeinstruction,