WPC1* _~ǧ1#LӾe}B'',[DjNS )NpF(3)6seKY*#;Jd3Y&` qҏ,SzJ&ЯȓK:CND5Y8*8ۑpUpZ6!8SVU-*YaS?_=Z7g̖>աX ,#/%̟.qǩv3+; {P_, O.^.9>B)e㶘$e e!T}dv(0l>+X7QӀwH%FJpOH*6QVʭD}yrT"M=l6N tVVW-_l0HK{9L6~pUL*xm;~c1*guSRiŸ́Gby?wKsebセ/9^;o:\UrEc#! UN %1 0(7w@_4U&4 1 1u 0cB 1 0ct 0 c 1[: 0d 0A 1s: 0c 0T 0wd  0 1uy 07b 04dP 05h 02|UL 0d 0XNHUB BL C fa D3%X mZ 0Dqf{a} B(*U<U0 0D AQ 0DH 0D D/ B D3 BOl' ! DC" B-"2#C# $$ 0D% B*[%^ %%%%%N%%%%%%%%%M%% D3% i7&&&KO& 0je'''z'z'z' C'''f'a'''''''''''' D3( B*( C$)$)$)$)$)$)$)$)$)$)$)$)$)$)$)$)$)$) C>)>) 0LX)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X)X) DK) BB)Epson Stylus COLOR ESC/P 20&w0 Z 6Times New Roman RegularX($ /{/{//{//{/S/E//g}%3|x/ `CG TimesW1C4O37;CO_s11.11.1.11.1.1.11.1.1.1.11.1.1.1.1.11.1.1.1.1.1.11.1.1.1.1.1.1.1##37=CIQYag1.a.i.(1)(a)(i)1)a)(;3$2#  0  .3  0  C37;CO_s11.01.0.01.0.0.01.0.0.0.01.0.0.0.0.01.0.0.0.0.0.01.0.0.0.0.0.0.0(;3$2#  a  )3  0  (;s$2#  a  .3  0  y379;=?AGM1.a.1)(Wf|$ @*  A (Vm$0   #37=CIOW_eI.A.1.a.(1)(a)i)a)(;3$2#  0  .3  0  (,!$0  0` (#(#  (O;$0  2#  a  .3  0` (#(#(b$0  0` (#(#2#   .3  0 ` (#` (#'#137=CIQYag1.a.i.(1)(a)(i)1)a)(:[$2#   .3  0  (<w$$2#(  0  )3  0  (@$0  0` (#(#0 ` (#` (# ` (T$0  0` (#(#0 ` (#` (#0 (# (# &w0 Z 4Times New Roman Italic(<,$2#(  a  )3  0  >$"Small Circle"0  ) `*Times New RomanTT d d\TABLE A""""'dxdBU(3$ !  ݁Level 1Level 2Level 3Level 4Level 5Table_A&0 d dLevel 1Level 2Level 3Level 4Level 5\  `$Times NewRoman2 `CG TimesLevel 1Level 2Level 3Level 4Level 5(3$ !  14256C<< C(3$ !  ($$   1  77' dxdP Pd9999'dxd  d 4"  EXAMPLE:   Acidcausesskinburns.(active)   X   Burnstotheskinarecausedbyacid.(passive)  4"  EXAMPLE:   &/ %XXInhalationofthismaterialcancause:  $,XX$ ,  X Noseandthroatirritation   ,  X Shortnessofbreath   ,  X Coughing tt  ,  X SneezingLevel 1Level 2Level 3Level 4Level 5>>>>7Hairlinedxd)Hairline d ((G(хLevel 1Level 2Level 3Level 4Level 5    C'  _SXXSECONDDRAFTREPORTONHAZARDCOMMUNICATION,UniversityofMarylandMedicalSchoolm!m!E(#PAGE  1  #XRXS+#(3$ !  &p Pd dwU% C )&  Needtoincludethegradelevelinhere.SDDDD'dxd&"#-%1, 2, 3,Level 1Level 2Level 3Level 4Level 54#?2Quick 1.Y  .0 Y dTABLE Bf  )&  _Thesearerecommendationsnotconclusions.Theyaremoreeditorialthanderivedfromtheliteraturesearch.Additionally,theytendtobefocusedontheU.S.Weshouldrestructurethistoreflectinternationalconcerns.qqqq'dxd&p d Pdddd>b$Large Circle0 ?Thin/Thick 2dxd>Thin/Thick 2[ -d d !  _C 4OS# CXX B  55TRX3' Letter3'3' Letter3'T/rr6(#///IOMC/////#XRXn#////&7%XXR//_IOMC_/ILO/HC2/99._Inf_.2//#XRX%&7#XXR/@@*/0 @ 0@p(#(#0@p(#p(#0@ (#(#//0@x (# (#/  #XRX`#XXR  SECONDDRAFTDOCUMENT#XRXB#XXR   a#XRXѥ#  L3MXXRHazardCommunication:#XRXML3# ! !XXR =  #XRX ! !q#XXRAReviewoftheScienceUnderpinning ;  theArtofCommunicationforHealthandSafety#XRX# ! !XXRԀ#XRX ! !q# L  $ XXRSubmittedto_ToxaChemica_,International ] inasubcontracttotheOccupationalSafetyandHealthAdministration#XRX $#$ XXRContractnumber:J9F50050 | TaskOrderNo.2,OptionYear1#XRX $#$ XXR @ PreparedbyBarbara_Sattler_,R.N.,Dr.P.H., b Bruce_Lippy_,_CIH_,CSP,andTyroneG.Jordan,B.S.#XRX $J#$ XXRTheEnvironmentalHealthEducationCenter # TheUniversityofMarylandMedicalSchool#XRX $k # $ May23,1997XXR #XRXH #  )%! @[[[[@@" TABLEOFCONTENTS      Vm  VmB 0     10(#(#INTRODUCTION#""J(#. (#(##4VmB ^ ݌  Ќ  20  SEARCHMETHODSANDINFORMATIONSOURCES#""J(#.!!5(#(##6  30  REVIEWOFLITERATURE#`"`"I(#.''#(#(##10      0  0` (#(#0 ` (#` (# `   3.1.10 (# (#SUMMARYOFRESEARCHWITHRECOMMENDATIONS#`"`"I(#.bbA (# (##10 ݌   Ќ     0  0` (#(#0 ` (#` (# `   3.1.20 (# (#INTRODUCTION#`"`"I(#.% (# (##12 ݌   Ќ    0  0` (#(#0 ` (#` (# `   3.1.30 (# (#COMPREHENSIBILITY#`"`"I(#.* (# (##13 ݌  z Ќ    0  0` (#(#0 ` (#` (#0 (# (#   3.1.3.10(#(#UseofSymbolsand_Pictograms_Ԁ#`"`"I(#.EE<(#(##14݌  g Ќ    90  0` (#(#0 ` (#` (#0 (# (#   3.1.3.20(#(#Shapeoflabelsandwarnings#`"`"I(#.==:(#(##179T݌ lT  Ќ    \0  0` (#(#0 ` (#` (#0 (# (#   3.1.3.30(#(#UseofMultipleMedia#`"`"I(#.<<3(#(##18\w݌ YA  Ќ    x0  0` (#(#0 ` (#` (# `   3.1.40 (# (#READABILITY#`"`"I(#.$ (# (##19x݌ F.  Ќ    t0  0` (#(#0 ` (#` (#0 (# (#   3.1.4.10(#(#Legibility#`"`"I(#.((#(##19t݌ 3  Ќ    0  0` (#(#0 ` (#` (#0 (# (#   3.1.4.20(#(#Readinglevelevaluations#`"`"I(#.))7(#(##20݌    Ќ    0  0` (#(#0 ` (#` (#0 (# (#   3.1.4.30(#(#TypeFontsandTypeSize#`"`"I(#.336(#(##20݌    Ќ    0  0` (#(#0 ` (#` (# `   3.1.50 (# (#PLACEMENTOFINFORMATION#`"`"I(#.1 (# (##21݌   Ќ    0  0` (#(#0 ` (#` (# `   3.1.60 (# (#AUDIENCEVARIABLES#`"`"I(#.$$+ (# (##23݌  Ќ    0  0` (#(#0 ` (#` (#0 (# (#   3.1.6.10(#(#Generalconsiderations#`"`"I(#.6(#(##23݌  Ќ    0  0` (#(#0 ` (#` (#0 (# (#   3.1.6.20(#(#Gender#`"`"I(#.[[$(#(##24 ݌  Ќ    0  0` (#(#0 ` (#` (#0 (# (#   3.1.6.30(#(#Age#`"`"I(#.WW!(#(##24݌  Ќ    0  0` (#(#0 ` (#` (#0 (# (#   3.1.6.40(#(#Hazardandriskperception#`"`"I(#.8(#(##25!݌  Ќ    '0  0` (#(#0 ` (#` (#0 (# (#   3.1.6.50(#(#Stress#`"`"I(#.$(#(##30'B݌ p Ќ    40  0` (#(#0 ` (#` (#0 (# (#   3.1.6.60(#(#Productfamiliarity#`"`"I(#.1(#(##304O݌ u] Ќ    N 0  0` (#(#0 ` (#` (#0 (# (#   3.1.6.70(#(#Color#`"`"I(#.#(#(##31N i ݌ bJ Ќ    Z!0  0` (#(#0 ` (#` (#0 (# (#   3.1.6.80(#(#CostofCompliance#`"`"I(#.XX0(#(##33Z!u!݌ O7 Ќ    s"0  0` (#(#0 ` (#` (# `   3.1.70 (# (#MEASUREMENTANDPROTOCOLS#`"`"I(#.002 (# (##39s""݌ <$ Ќ    }#0  0` (#(#0 ` (#` (# `   3.1.80 (# (#STANDARDPHRASES#`"`"I(#.) (# (##40}##݌ ) Ќ    ~$0  0` (#(#0 ` (#` (#0 (# (#   3.1.8.10(#(#Signalwords#`"`"I(#.((*(#(##40~$$݌  Ќ    %0  0` (#(#0 ` (#` (#0 (# (#   3.1.8.20(#(#Standardizationoflabelformat#`"`"I(#.=(#(##43%%݌   Ќ    &0  0` (#(#0 ` (#` (# `   3.1.90 (# (#ADDITIONALRESEARCHNEEDS#`"`"I(#.gg2 (# (##44&&݌   Ќ  !  !'0  0` (#(#     ` 3.20 ` (#` (#TRAINING#`"`"I(#.ll! (# (##45!''݌ ! Ќ    (0  0` (#(#0 ` (#` (# `   3.2.10 (# (#SUMMARYOFFINDINGSANDRECOMMENDATIONS#`"`"I(#.@ (# (##45((݌ " Ќ    )0  0` (#(#0 ` (#` (# `   3.2.20 (# (#GENERALCONSIDERATIONS#`"`"I(#.../ (# (##46))݌ #  Ќ    *0  0` (#(#0 ` (#` (# `   3.2.30 (# (#_NIOSH_ԀLITERATUREREVIEW#`"`"I(#.880 (# (##46**݌ $! Ќ    +0  0` (#(#0 ` (#` (# `   3.2.40 (# (#RIGHTTOUNDERSTAND#`"`"I(#., (# (##48+,݌ %y" Ќ    ,0  0` (#(#0 ` (#` (# `   3.2.50 (# (#ADDITIONALRESEARCHNEEDS#`"`"I(#.gg2 (# (##50,-݌ ~&f # Ќ    .0  0` (#(#0 ` (#` (# `   3.3.10 (# (#SUMMARYOFFINDINGSANDRECOMMENDATIONS#`"`"I(#.@ (# (##51. .݌ k'S!$ Ќ    /0  0` (#(#0 ` (#` (# `   3.3.30 (# (#HAZARDCOMMUNICATIONASAWHOLEANDDYNAMIC X(@"% PROCESS#`"`"I(#. (# (##62/8/݌ E)-#& Ќ    U00  0` (#(#0 ` (#` (# `   3.3.40 (# (#ADDITIONALRESEARCHNEEDS#`"`"I(#.gg2 (# (##67U0p0݌ 2*$' Ќ  40  CONCLUSIONS#`"`"I(#.i i (#(##67 +%( 50  BIBLIOGRAPHY#`"`"I(#. (#(##69  ,%)  6   ,&*  ?> /A.    3    3]224O  1  3  0    INTRODUCTION P2  3]22݌(#(# Ќ      ThisreportwascommissionedbytheUnitedStatesOccupationalSafetyandHealthAdministrationtoreviewthestateofscientificinquirysupportingourknowledgeregardingkeyelementsofchemicalhazardcommunicationprograms:labeling,warnings,materialsafetydatasheets,andworkertraining.Thisendeavorsupportstheinternationalefforttoharmonizelaws,regulations,andconsensusstandardsaffectingthewaysinwhichinformationabouthazardouschemicalsiscommunicated.Theinternationaleffortcanbedividedintothreemajorfunctions:classifyinghealthandenvironmentalhazards,classifyingphysicalhazards;andcommunicatinghazardinformation.Thislastcomponentinvolvesthedeterminationofwhatinformationwillbecommunicatedtousersregardingthehazardsandappropriateprotectivemeasures,aswellasthewayinwhichitwillbetransmitted,i.e.throughsymbols,labels,standardphrases,andtraining.  Internationally,sixorganizationstheWorldHeathOrganization,theInternational_Labour_Organization,theUnitedNationsEnvironmentalProgram,theFAO,_UNIDO_,andOECDparticipateintheInterOrganization_Programme_ԀfortheSoundManagementofChemicals(_IOMC_).The_IOMC_Ԁhascompiledabibliographyofreferencesandhasadministeredasurveysenttomembercountriesaskingforinformationabouttheirrespectiverequirementsonhazardcommunication.Additionally,thecountriesrespondingtothesurveywereaskedtosubmitanystudiesandsupportingdocumentationfortheircountryspolicies,lawsandregulations.TheresultingdraftreportwasproducedbytheWorkingGrouponHarmonizationofChemicalClassificationSystemswhichwasestablishedbytheCoordinatingGroupofthe_IOMC_.ThefulltextofthisimportantreportiscontainedAppendixAandisreferencedthroughoutthisdocumentasthe_IOMC_Ԁdraftreport.Thefollowingkeychartistakendirectlyfromthereport.*}c"dVddd Xdd Xdd X(#(#},zdd ,dd ,~dd ,dd ,dd ,dd ,dd ,dd ,dd ,Zdd , dd +   <$  -#!"# -SXXRAustralia#XRXS <# -#!"# -SXXRCanada#XRXS<# -#!"# -SXXRJapan -#!"# -Korea -#!"# -Mexico -#!"# -Portugal -#! "# -Sweden -#!!"# -UK -#!""# -USA -#!#"# -Zimbabwe 2(&$"  # 2 #XRXS<#YXXRLabels#XRXY>#SXXR 'x`% '#XRXS?#YXXRyes#XRXY?#SXXR 'x`&" '#XRXS?#YXXRyes#XRXY8@#SXXR 'x`'" '#XRXS}@#YXXRyes#XRXY@#SXXR 'x`(" '#XRXS+A#YXXRyes#XRXYA#SXXR 'x`)" '#XRXSA#YXXRyes#XRXYBB#SXXR 'x`*" '#XRXSB#YXXRyes#XRXYB#SXXR 'x`+" '#XRXS5C#YXXRyes#XRXYC#SXXR 'x`," '#XRXSC#YXXRyes#XRXYLD#SXXR 'x`-" '#XRXSD#YXXRyes#XRXYD#SXXR x`." #XRXS?E#YXXR?#XRXYE#SXXR  x`/ #XRXSE#YXXRText#XRXY9F#SXXR ' 0 '#XRXSF#YXXRyes#XRXYF#SXXR ' 1" '#XRXS-G#YXXRyes#XRXYG#SXXR ' 2" '#XRXSG#YXXRyes#XRXYDH#SXXRԀ ' 3" '#XRXSH#YXXRyes#XRXYH#SXXR ' 4" '#XRXS8I#YXXRyes#XRXYI#SXXR ' 5" '#XRXSI#YXXRyes#XRXYOJ#SXXR ' 6" '#XRXSJ#YXXRyes#XRXYJ#SXXR ' 7" '#XRXSBK#YXXRyes#XRXYK#SXXR ' 8" '#XRXSK#YXXRyes#XRXYYL#SXXR  9" #XRXSL#YXXR?#XRXYL#SXXR   : #XRXSAM#YXXRSymbols#XRXYM#SXXR *"; #XRXSM#YXXR  no#XRXYAN#SXXR *"<  `N#XRXSN#YXXR  yes#XRXYO#SXXR *"=    #XRXSRO#YXXRno *">    yes '*"?  'yes '*"@" 'yes '*"A" 'yes '*"B" 'yes '*"C" 'no '*"D" 'yes *"E" Colors #kF   no(no) H$0H    no(no) H$0J    no(no) H$0L    yes(yes) H$0N  Q  yes(yes) 'H$0P  'no(yes) 'H$0R" 'no(yes) 'H$0T" 'no(yes) H$0V" R  no(no) 'H$0X  '?(yes) H$0Z" Format '%[ 'no '%\" 'yes '%]" 'no '%^" 'yes '%_" 'yes '%`" 'yes '%a" 'yes '%b" 'yes '%c" 'no '%d" '? '%e" '#XRXYO#SXXR Safetydata & f sheets #XRXS=U#YXXR ''!g" 'yes '& h" 'yes '& i" 'yes '& j" 'yes '& k" '? '& l" 'yes '& m" 'yes '& n" 'yes '& o" 'yes '& p" '? '& q" 'Headings '("r" 'yes '("s" 'yes '("t" 'yes '("u" '? '("v" '? '("w" 'yes '("x" 'yes '("y" 'yes '("z" 'yes '("{" '? '("|" 'Symbols '!* $}" 'no '!* $~" 'no '!* $" 'no '!* $" '? '!* $" '? '!* $" 'no '!* $" 'no '!* $" 'no '!* $" 'no '!* $" '? '!* $" ' Training 'z+b%" 'yes 'z+b%" 'yes 'z+b%" 'yes 'z+b%" '? 'z+b%" '? 'z+b%" '? 'z+b%" 'yes 'z+b%" 'no 'z+b%" 'yes 'z+b%" '?"z+b%" "#XRXYU#  \-D' A. A A  Legend YXXR / R  Labels:0  0` (#(#iflabelsarerequiredbyregulationsorrecommended.` (#` (# Text:0  0` (#(#ifanyspecificwritteninformationisrequiredorrecommendedonthelabel.` (#` (# Symbols:0 ` ifanyspecificsymbolsarerequiredorrecommendedonthelabel.v` (#` (# Colors:/0  //0` (#(#/ifthelabelisrequiredorrecommendedtohavespecificcolors;parenthesesindicateifsymbolsare j ; requiredorrecommendedtohavespecificcolors. ` (#` (# Format:/0 ` /ifthelabelisrequiredorrecommendedtohaveaspecificsize,shape,ordesign. ` (#` (# SafetyDataSheets:ifdatasheetsarerequiredbyregulationorrecommendedHeadings: / / / / /0 `  /ifstandardizedheadingsarerequiredorrecommendedonthedatasheet.~ O` (#` (# Symbols: /0 `  /ifspecificsymbolsarerequiredorrecommendedonthedatasheet.C  ` (#` (# Training: /0 `  /iftrainingisrequiredorrecommended. ` (#` (# ?/0  //0` (#(#/meanstherewasnoindicationinthedocumentsreceived. ` (#` (# e~#XRXYx]#  Thereisastrongconsensusthatafullyharmonizedhazardcommunicationsystemmusttake W intoaccountthescientificfindingsoncomprehensibility,readability,andotherhumanfactorsregardingtheuseoflabels,warningplacards,andsafetydatasheets.Further,thiseffortmustbebuiltonpreviouswork.Itisimpracticalandilladvisedtoignorethefactthatthereareexistingsystems,severalusedinternationally,thatshouldberevisedtomakethemmoreeffective.  Anextensiveliteraturesearchwasconductedonhazardcommunicationandtheinformationissummarizedwithinthisreport.Notallcitationsenteredinthebibliographyarecitedwithinthebodyofthereport.SeveralU.S.agencieshavedonesimilarreviewsofliteratureregardingeffectivenessoflabeling,warningsymbols,andhazardouschemicaltraining:theEnvironmentalProtectionAgency(EPA),theConsumerProductSafetyCommission(_CPSC_),theNationalInstituteforOccupationalSafetyandHealth(_NIOSH_),andtheNationalInstituteforEnvironmentalHealthScience(_NIEHS_).Thereisstronginternationalinterestinthecomprehensibilityofinformation.Inadditiontosummarizingtheliterature,deficienciesinourknowledgehavebeenidentifiedandrecommendationsforfurtherstudieshavebeensuggested.    3    3!i24O  2  3  0    SEARCHMETHODSANDINFORMATIONSOURCES i 3!iLi݌' (#(# Ќ    InDecember1996,theILOsWorkingGrouponHarmonizationofChemicalHazardCommunicationissuedadraftdocumententitled Reportontheresponsestothecallonchemical !# hazardcommunication.ThecallsforinformationwereactuallysentinMayandJuly1995toabout " $ 270relevantnationalandinternationalinstitutionswithaformalquestionnairegoingtoafeworganizationsselectedfromthislargerbody.TherequestswentoutthroughtheILOsInternationalOccupationalSafetyandHealthInformation_Centre_.Backgroundmaterialsonhazardcommunicationwererequested,particularlyscientificstudies,withanemphasisoncomprehensibilitystudiesandproceduresusedtoimplementhazardcommunicationsystemsatthenationallevelaswellasbyindustry.Theinformationwascollectedasbackgroundinformationforcreatingaharmonizedchemicalhazardcommunicationsystems.  Theworkinharmonizationisbasedonanumberofinternationalresolutions,conventions,andreports.InJune1992,theUnitedNationsConferenceonEnvironmentandDevelopment(_UNCED_)stipulatedthat agloballyharmonizedhazardclassificationandcompatiblelabeling  -*/ systemincludingmaterialsafetydatasheetsandeasilyunderstandablesymbols,shouldbeavailable,iffeasible,bytheyear2000.PartBofChapter19of_UNCED_sAgenda21isdevotedtoharmonizationofclassificationandlabeling.////  TheWorkingGrouponHarmonizationofHazardCommunicationaddressesissuesregardinghazardsymbols,colors,andwritteninformationusedonlabels;thepreparationofchemicalsafetydatasheetsandinstructions;thecomprehensibilityofprecautionarystatementsusedonbothlabelsandinchemicalsafetydatasheets;andtrainingrelatedtotheseareas.////  TheCoordinatingGroupfortheHarmonizationofChemicalClassificationSystemsbringstogetherexpertsrepresentingexistingnational,regionalandinternationalsystemsofclassificationandlabelingofchemicals,aswellasrepresentativesofinterestedinternationalorganizationsofsuppliers,employers,workers,consumersandenvironmentalgroups.Theobjectivesareafinalizedproposalforharmonizedclassificationcriteriaby1997andpossibleimplementationofagloballyharmonizedsystematthenationallevelbytheyear2000.////  Thefieldofoccupationalsafetyandhealthisbynature_multidisciplinary_Ԁandtheliteraturedescribingitsvariedworksmustbesearchedusingseveraltools.Severalbibliographicdatabasesweresearchedbackto1980.Nosingledatabaseidentifiedthenecessaryliteraturetoaccomplishthetaskathand.Thehealthandsafetydatabaseshadlittleinformationonlabeling.Thisinformationwasfoundinavarietyofdatabasesincluding_PsycLIT_Ԁwhichmoreextensivelyidentifiedthehumanfactorsandergonomicliterature.Thehealthandsafetydatabaseswerealsodeficientintheircoverageoftrainingefficacy;thatbodyofliteraturewasfoundineducationaldatabasesandtosomedegreeinthe_NIOSH_Ԁdatabase(_NIOSHTIC_).TheuseofInternetsearchingtoolsuncoveredadditionalinformationfromavarietyofsourcesincludinginternationalagencies,nationalagencies,universities,andprivateorganizations.Theliteraturecitedandreviewedwaslimitedsubstantiallytopeerrevieweddocuments,suchaspeerreviewedjournalsandgovernmentreports.Becausetheliteratureontheeffectivenessoflabeling,_MSDS_,andtrainingspecificallyinthecontextofworkplacehazardcommunicationwassurprisinglysparse,thesearchwasextendedtolookatotherapplicationsofhazardcommunication,suchastheuseoflabelsforhazardwarningsonconsumerproductsandinotherpublicsafetysituationssuchastransportationsafety.Inthisextendedsearch,asubstantialliteraturewasidentifiedregardingtheeffectivenessoflabelingandwarningsigns.  //  Additionally,agencystaffwereinterviewedinseveraloftheU.S.agenciesresponsibleforproductandchemicallabeling,includingtheEnvironmentalProtectionAgency,theConsumerProductSafetyCommission,andtheDepartmentofTransportation.StaffattheNationalFireProtectionAssociationwerealsocontacted.ChairmenandkeycommitteememberswerecontactedfromthetwomainAmericanNationalStandardsInstitute(ANSI)standardscommitteesonlabelingand_MSDS_.MembersoftheSocietyforChemicalHazardCommunicationwerealsocontactedasameansofidentifyingandlocatingmaterialsandstudiesthatmayhavebeendonewithincompaniesortradeassociationsandthereforenotretrievablethroughbibliographicsearches.Somemembersofthe_HazCom_ԀWorkgroupwithintheNationalAdvisoryCommitteeonOccupationalSafetyandHealthwerealsocontactedforthesamereason. -&* Ї  OSHAprovidedabibliographicsearchthathadbeendonebytheInternationalOccupationalSafetyandHealthInformation//_Centre_//attheILO.ApproximatelyhalfofthedocumentslistedwerewritteninlanguagesotherthanEnglish.Themajorityofthedocumentslistedinthisbibliographyaregovernmentalreports,guidance,andinformationdocuments.Fewofthedocumentsarearticlesorbooksdescribingresearchontheeffectivenessofhazardcommunication..////  Asubstantialbibliographicsearchhadbeendoneasthefirstphaseofacurrent,U.S.,multiagencyprojectentitledthe ConsumerLabelingInitiative.Thisprojecthasthestatedgoalofmakingthehealthandsafetyinformationonlabelseasierforconsumerstounderstand.Thisbibliographyprovidedarichsourceofliteratureincludingagencyreportsandresearchonlabeling.Thesecondphaseofthisprojectwillcontinuein1997.TheConsumerProductsSafetyCommissionreceivedareportinMay,1995entitled, ProductLabelingGuidewhichcontainedanexcellentliteraturereview.ThereportwaspreparedbyBariM.//_Kotwal_//andNeilD.Lernerof//_COMSIS_//Corporationandhasbeenquitevaluableforthispresentreport.////  TheSocietyforChemicalHazardCommunicationproducedacompendiumofcitationsentitled, PublishedResearchontheComprehensibilityofChemicalHazardWarningLabelsandMaterialSafetyDataSheets.Althoughthislistedanumberofarticlesonlabeling,thevastmajorityofarticlesdescribedlabelsinnonworkplacesettings,suchasonconsumerproductsandintransportationcontexts.Further,onlyonearticlelistedreferredtoMaterialSafetyDataSheets.////  In1995,aliteraturereviewentitled AssessingOccupationalSafetyandHealthTrainingwaspublishedbythe//_NIOSH_//thatsoughtevidencefromtheliteraturebearingontwoquestions:AreOSHAtrainingrequirementseffectiveinreducingworkplaceinjuryandillnesses?Doestheavailableevidenceshowcertaintrainingfactorsorpracticesasbeingmoreimportantthanothersinhavingapositiveeffectonillnessandinjuryprevention?Additionally,asetofpaperswascollectedfora1996workshopsponsoredbytheNationalInstituteforEnvironmentalHealthScience,entitled MeasuringandEvaluatingtheOutcomesofTraining.////  Alloftheabovementionedsearches,reviews,andlistingsareincludedinthisreporteitherasindependentdocumentsorsubsumedintothecomprehensivebibliography.̀Thefollowingdatabasesweresearchedforthisreportusingthekeywordsidentifiedforeachdatabase://1.//DIALOG,ABI/INFORMand//_COMPENDEX_//  KEYWORDS:0  riskmanagement,safetymanagement,humanfactorsand 'T!$ communication,behaviorandcommunication,compliance/noncompliance,chemical,toxic,worker,employee,occupationalhealthoccupationalinjury,health,injury (#(# ////// -&* _2.//MEDLINE//(totalofthreesearches)  KEYWORDS:0  materialsafetydatasheets,hazardcommunication,risk  communication,hazardoussubstances,hazardouschemicals,international,productlabeling,label* (#(# 3.IPA(InternationalPharmaceuticalAbstracts)  KEYWORDS:0  packageinserts,labels,safety,hazard,caution,compliance/non   compliance,hazardcommunication,riskcommunication (#(# 4.//_PsychLIT_//  KEYWORDS:0  decision*,workingconditions,compliance/noncompliance, p  workplace,employ*,safety,health,warn*,humanfactors,behavior,materialsafetydatasheet (#(# 5.//_NIOSHTIC_//  KEYWORDS;0  training,materialsafetydatasheets,_MSDS_,hazardcommunication_0 (#(# *Denotesanyvariationontherootword.    3    324O  3  3  0    REVIEWOFLITERATURE  3=݌/(#(# Ќ   s   s24O  3  .  1  3  0    LABELSANDWARNINGS s/݌(#(# Ќ    [   [24O  3  .  1  .  1  3  0    SUMMARYOFRESEARCHWITHRECOMMENDATIONS  [3݌(#(# Ќ    Thefollowinginformationrepresentsasummaryofthegeneralareasofconsensusintheliteratureonlabelsandwarnings.Thesummaryalsoincludeswhattheauthorsofthisreportconsiderkeyfindingsandrecommendations.Z?  Z?(Y1  .0    Itisdifficulttofindcomprehensive,usableanddefinitiveguidanceforthedesignor ` evaluationoflabels,inspiteoftheexistenceofseverallabelstandardssuchasANSI//Z535//standardsandthe//_ISO_//standards.Z?(CYی (#(# Ќ  Z?  Z?Y2  .0    Labeleffectivenessisoftenjudgedonthebasisofsubjectiveopinion,particularlywhat _!0 wouldbedoneuponencounteringthelabel.Therelativevalidityandultimatesafetyconsequencesoflabelingalternativesarepoorlyunderstood,however.Z?+Yی (#(# Ќ  Z?  Z?̟Y3  .0    Muchoftheliteraturehasalsobeencharacterizedbytheuseofcollegestudentsasresearch /$  subjects,ratherthanworkers,thekeytargetaudience.Z?̟Yی (#(# Ќ  Z?  Z?(Y4  .0    Perceivedhazard,familiaritywithaproduct,andgenderinfluencetheconsumer'sdecision &" tolookforawarningmessageonthelabelsofpotentiallyhazardoushouseholdproducts.Z?(CYی (#(# Ќ  Z?  Z?Y5  .0    Warningslaidoutinoutlineformandorganizedbytypeofhazardwererankedashaving '!$ greatereyeappeal,easiertoprocess,andmoreeffectivethanalternativeapproachessuchasparagraphlayouts.Z?Yی (#(# Ќ  Z?  Z?5Y6  .0    Educationaleffortsareessentialifsymbolsorcolorsaretocommunicateeffectively.Even *$' brieftrainingaslittleasgivingthepictorialsverbalmeaningoncecanhavealargeimpactinfacilitatingcomprehensionforpictorialsthatwouldotherwisenotbeunderstoodbymanypeople.Z?5PYی-`'*(#(# Ќ  Z?  Z?*Y7  .0    FlameandpoisonwarningsymbolsprescribedbytheCanadiangovernmentarewell / understoodgenericallybutpeoplehavedifficultyinferringthespecificsafetyprecautionsmostnecessaryforaparticularproduct.Z?*EYی (#(# Ќ  Z?  Z?ŧY8  .0    Theadditionofpictographstoaverbalwarningwillincreasecomplianceratesbutatleast  onekeystudyshowedthatenclosureshapesmadenodifferenceincompliancerates.Z?ŧYی (#(# Ќ  Z?  Z?:Y9  .0    Conductingbehavioralstudiesispreferabletorelyingonpreferencedatabutisrarely   conducted.Z?:UY ی (#(# Ќ  Z?  Z?cY10  .0    StudiesoftheANSI//Z535.3//1991protocolfortestingfoundthatthestandardslash   recommendedbyANSIwasnotaseffectiveasan Xforwarningconsumersofprohibiteduses.Z?c~Y ی (#(# Ќ  Z?  Z?Y11  .0    Warningscontainingapictorial,color,oraniconelicitedsignificantlyfasterresponsetimes `  amongsubjectsthanwarningswithoutthem.However,theadditionofaborderdidnotimproveresponsetimes.Z?Y ی (#(# Ќ  Z?  Z?Y12  .0    Theshapeofalabelmayenhancetheconspicuousnessofawarning,evenifthetargeteduser _0  isnotactivelysearchingforsafetyinformation.Thismaybebecausethepublicmaybecomemoresensitivetocertainshapestimuliwhichtheyhavecometoassociatewithdangermuchastheydowithcolor.Z?Y ی (#(# Ќ  Z?  Z?~Y13  .0    Evenwithreasonableattemptstomandatelegibilitytherehavebeenanumberoffactorsnot  specifiedthatcanhavesignificanteffects.Z?~Y ی (#(# Ќ  Z?  Z?аY14  .0    X]XXXbOneofthekeyvariablesonwarningeffectivenessiswheretheinformationisplaced.#XbXXX]0#Z?аYی(#(# Ќ  Z?  Z?$Y15  .0    Warningsprintedhorizontallywerefoundmorequicklythanwarningsprintedvertically.Z?$?Yی(#(# Ќ  Z?  Z?7Y16  .0    Ifsafetyinformationisperceivedasfamiliarorredundant,itmaysimplybedroppedfrom  shorttermmemory,andhavenofurthereffectonbehavior.Alternatively,iftheinformationisperceivedasinconsistentwithexistingknowledgeandbeliefs,itmayberejectedasnotcredible.Eveniftheinformationispreviouslyunknownandisacceptedastrue,therecipientmaybeinsufficientlymotivatedtoalterbehavior.Z?7RYی (#(# Ќ  Z?  Z?Y17  .0    Widelyencounteredguidelinesthatwarningsshouldcomprisefourelementsastandard b signalword,statementsofthehazard,thepotentialconsequences,andhowtoavoidthehazardareprimarilybasedonpreferencestudiesandmaybeinerrortosuggestthatallofthisinformationmustbeprovidedinordertoelicittheappropriateresponse.Z?Yی (#(# Ќ  Z?  Z?˷Y18  .0    Womenreportagreaterlikelihoodthanmentolookforandreadwarnings.Theyarealso Q"" morelikelytocomplywithwarnings.Z?˷Yی (#(# Ќ  Z?  Z?Y19  .0    Itmaybenecessarytoapplyrelativelystrongsignalwordsforolderusers:theyexhibita 1$  lowerlevelofcomprehension.Thoseover40aremorelikelythanyoungersubjectstotakeprecautionsinresponsetowarningstheyunderstand,however.Z?*Yی (#(# Ќ  Z?  Z?̺Y20  .0    Perceivedseverityofconsequenceshasastronginfluenceonbehavioralintentions.Z?̺Yی' #(#(# Ќ  Z?  Z?ܻY21  .0    Higherstressproducesignificantlylowercompliance.Z?ܻYی'!$(#(# Ќ  Z?  Z?μY22  .0    Hazardinformationcommunicatedbydifferentcolorsfollowedaconsistentpatternacross ("% languagegroups.Redresultedinthehighesthazardratingsfollowedbyorange,yellow,blue,green,andwhite,respectively.Z?μYی (#(# Ќ  Z?  Z?mY23  .0    TheimportancethatANSIandtheU.S.militaryattachestosignalwordsmaynotbeshared +%( bythepopulationatlarge.MoststudiesindicatedthatDANGERconnotedgreaterstrengththanWARNINGandCAUTION.Butgenerallytheresultsfailedtoshowadifference -b'* _betweenWARNINGandCAUTION.Z?mYی (#(# Ќ    [   [24O  3  .  1  .  2  3  0    INTRODUCTION } [݌(#(# Ќ    Oneofthemostcomprehensivereviewofliteratureonlabelsandwarningswasperformedby//_COMSIS_//CorporationundercontracttotheU.S.ConsumerProductSafetyCommission(//_Kotwal_//andLerner,1995).Theauthorsdescribedthestateofknowledgeonlabelsandwarningswiththefollowingexcellentsummary:0  Althoughwarninglabeldesignmaysuperficiallyappeartobeasimplematter,properdesignanduseis,infact,rathercomplex.Itisclearthatinappropriatelydesignedorusedwarningssimplywillnotbeeffective,byanystandardofevaluation.Yet,itisdifficulttofindcomprehensive,usableanddefinitiveguidanceforthedesignorevaluationoflabels,inspiteofthefactthattherearenumerouslabelstandards(e.g.,ANSI//Z535//standards,//_ISO_//standard),labeldesignguides(e.g.,Westinghouseand//_FMC_//handbooks),generalhumanfactorsguidanceforwarningsandlabels,andvariousliteraturereviewarticles...Unfortunately,thefindingsoftheresearchliteratureonlabelsareoftencontradictoryandfrequentlyweakintermsoffacevalidityorexperimentalprocedures.Labeleffectivenessisoftenjudgedonthebasisofsubjectiveopinionorpeople'sreportsofwhattheythinktheywoulddoiftheyencounteredthelabel.However,therelativevalidityandultimatesafetyconsequencesoflabelingalternativesispoorlyunderstood. (#(# 0  Theliteraturedoescontainanumberofexperimentalstudiesthatdiddirectlymeasuresomebehavior,suchasreadingalabelorcomplyingwithsomeaction,althoughthesestudieshaveusuallybeenrathernarrowinapproach.Furthermore,therehavebeenmanyexamplesofpoorresearchdesign,orinappropriateconclusionsdrawnfromdata.Muchoftheliteraturehasalsobeencharacterizedbytheuseofsamplesofconvenienceforthepopulationofresearchsubjects.Veryoftenthesubjectsarecomprisedofcollegestudents.Thisis,atbest,anarrowsampleand,quiteoften,aninappropriateone,asthesepeoplemaydifferfromtypicalusersoftheproductintermsofproductknowledge,hazardperception,perceptualabilities,orsafetymotivation.Also,theresearchliteraturehassometimesbeencharacterizedbyasortofnaiveabsolutism.Forexample,astudymayaskthequestionwhetheritismosteffectivetopresentthemessageelementofthehazard,therequiredaction,theinjuryoutcome,orthelevelofdanger,orsomecombinationofthese.Whilesuchresearchiscertainlyuseful,whendesigninganactuallabel,thequestionverylikelybecomeswhatinformationismosteffectiveforthisapplication(consideringthehazard,publicunderstandingofthehazardandrelatedactions,spaceconsiderations,theobviousnessofthehazardontheproduct,competitionwithothermessages,andmanyotherfactors).Thus,thelimitationsofexistingresearchmakeitdifficultforadesignertofinddefinitiveguidanceforpreparinganeffectivelabelorforananalysttomakeconsistentanddefensibleevaluationsofproductlabelingpractices. (#(# // //// -`'*    [   [ 24O  3  .  1  .  3  3  0    COMPREHENSIBILITY  [ 5݌/(#(# Ќ    Comprehensibilityreferstotheabilityoftheindividualreadingalabel,warning,ormaterialsafetydatasheettounderstandtheinformationsufficientlytotakethedesiredaction.Comprehensibilityisdifferentfromreadabilitybecausethelatterissimplyameasureofthegradelevelofthewrittenmaterialwhiletheformerisameasureofhowwellthereceiveroftheinformationunderstoodit.Awarningaboutincompatiblechemicalsmaybewrittenatthecorrectreadinglevelforaspecificaudiencebutmaydosuchapoorjobexplainingtheconceptofincompatibilitythatthewarningisntunderstandbymostoftheintendedaudience.Additionally,thesamewarningmaybehighlycomprehensibletoapopulationofchemicalworkersbutpoorlyunderstoodbyfirefighterswiththesameeducationallevelbutdifferentworkexperiences.Finally,achievingcomprehensibilitydoesnotensurethattheinformedindividualwilltaketheactionsprescribedinthewarningorlabel.Thisfinal,behavioralstepisaffectedbyacomplexmixofattitudes,experiences,motivations,andpotentialconsequencesthatarespecifictoeachindividualinaparticularsituation.////  X]XXXbTwoexperimentsinvolvingwarninglabelsonhouseholdproducts(Godfreyand//_Laughery_//, / //1993b//)indicatedthatsubjectsareabletodiscriminateamongproductsbasedonoverallhazard.Themorehazardoustheyperceivedaproduct,themorelikelyitwastheywouldlookforawarning.Further,femalesaremorelikelytolookforwarningsthanmales.Insummary,perceivedhazard,familiarity,andsexinfluencetheconsumer'sdecisiontolookforawarningmessageonthelabelsofpotentiallyhazardoushouseholdproducts.////  Otherresearcherslookedatthevariablesrelatedtopeople'sperceptionsofhazardforvarioustypesofhouseholdpestcontrolproducts(Silverand//_Wogalter_//,//1991a//).#XbXXX]#X]XXXbԀSeventycollege  agedstudents,20oldersubjects,andfourpesticideexpertsjudged26pestcontrolproductsonhazard.Resultsshowedthatfumigatorsand//_foggers_//weretheproductsperceivedasthemosthazardousfollowedbysprays,systems,andtraps,respectively.Althoughseveral//_misperceptions_//ofhazardforcertainproductcategorieswereevident,thejudgementsofthestudentsandolderadultswereconsistentwiththoseoftheexperts.Perceivedhazardwasalsofoundtobepositivelycorrelatedwithanumberofobjectivecharacteristicsoftheproductlabels,includingthequantityofchemicalingredients,effectivenessagainstpests,numberofwordsandsentencesonthelabel,readability,andthepresenceandlocationofcertainstatementsonthebacklabel.Thesefindingsuggestthatpeoplecandiscriminatethehazardlevelsofdifferenttypesofpestcontrolproducts,andthepresenceofvariouscuesonthelabelmayaidthisdetermination.//#XbXXX]#X]XXXb//  #XbXXX]G#//_Desaulniers_//(1980)examinedtheeffectsofwarninglayoutandsemanticorganizationonthe '!$ readabilityandrecallofwarninginformation.Ingeneral,warningsinoutlinelayoutandtypeofhazardorganizationwererankedashavinggreatereyeappeal,easiertoprocess,andmoreeffectivethanalternativeorganizationlayoutconditions(i.e.,paragraphlayouts).Theresultsindicatedthatwarningsinoutlinelayoutwerereadandcompliedwithbyalargerproportionofsubjectsthanwarningsinparagraphlayout.CollinsandLerner(//1982b//)alsonotedotherimportantcriteriainadditiontounderstandabilityinfiresafetyalertingmeasures.Variablessuchasvisualrange, -`'* _//detectability//insmoke,andlegibilityallwerecritical. X]XXXb /    w$   w$a24O  3  .  1  .  3  .  1  3  0 `   #XbXXX]#UseofSymbolsandPictograms T w$a݌` (#` (# Ќ    Therehasbeenagreatinterestforyearsindepictingcriticalsafetyandhealthinformationassuccinctlyaspossibleforimmediateresponsebyindividualsindangeroussituations.Thefocushasbeentoavoidthenumerousshortcomingsofwrittenlanguage,particularlythelengthoftimeittakestoreadawarningandtheproximitythereadermusthavewiththeinformationinordertoseeit.Additionally,theincreasingnumberofAmericanworkerswhodonotspeakEnglishasafirstlanguageorwhoareilliterateargueswellfortheuseofsymbolsand//pictograms//.Thisisalsoessentialforanyefforttowardsinternationalharmonization.////  Symbolsand//pictograms//arewidelyusedinwarninglabels.However,thistrend,atleastintheUnitedStates,isrelativelyrecent.Researchontheuseandeffectivenessofpictorial//signage//beganintheearly1970'sandcontinuedintotheearly1980's.Atthesametime,theuseofsymbolsburgeonedforapplicationssuchasthehighway,publicfacilities,workplacesigns,andproductlabels.Thesizablebodyofresearchonsymboleffectivenessfromthisdecadeofactivityprimarilyfocusedonsigns,forhighways,workplaces,andpublicfacilities,ratherthanlabels.Collins'1982reviewofliteraturethroughtheearly1980'sreportedthat"therehasbeenlittleifanyresearchontheeffectivenessofsymbolsforproductwarnings."Accordingto//Kotwal//andLerner(1995)althoughtherehasbeensubsequentresearchspecificallyonlabels,mostresearchstillhasbeenbasedonsigns;muchofcurrentpracticeisanextrapolationfromsignrelatedresearchorisbasedonlogicalargument.//X]XXXb//  Onestudyshowedcollegestudents16symbolsfromthe1981WestinghouseProductSafetyLabelHandbook.Foreachgraphic,thesubjectwasaskedtodescribethehazard.Recognitionratesrangedfromnearly100%toalmostcompletelyunrecognizable(//Laux//,MayerandThompson,1989).Resultsindicatedthattherearetwoquestionswhichmustbeaddressedinthedevelopmentofsymbolsandpictorials:1)willtheuserrecognizethehazard?and2)willtherecognitionofthehazardsuggestproperprecautionarybehavior?////  TheU.S.EPA(1986)receiveduniversalagreementfromtheiradvisorycommitteethateducationaleffortsareessentialifsymbolsorcolorsaretocommunicateeffectively.TheycitedtheexperienceoftheCanadiangovernmentwhichinstitutedasystemofsymbolstoconveydegreeofhazardinthemidseventies.Theysinceundertookaneducationaleffortwhichresultedin95%ofCanadiansunderstandingthemeaningofthesymbols.AsurveyconductedintheU.S.wherenosimilareducationalcampaignwasheldrevealedcomprehensionbelow25%.////  #XbXXX]#Becauseoftheirrelativelyuniversalinformationtransmissionpotential,pictorialshavebeen )#& suggestedascommonmeansofsafetycommunicationacrossheterogeneousgroupsofusersanduses(//Brelsford//,//Wogalter//,and//Scoggins//,1994).Theresearchersconductedastudyandfoundthateasypictorialswerecomprehended(bothinitiallyandfollowingtraining)betterthandifficultpictorials,althoughthelattershowedthemostdramaticincreaseinunderstandabilityaftertraining.Inaddition, -f'* theyfoundthatthesubstantialgainsmeasuredaftertrainingsuggestthatevenbrieftrainingaslittleasgivingthepictorialsverbalmeaningoncecanhavealargeimpactinfacilitatingcomprehensionforpictorialsthatwouldotherwisenotbeunderstoodbymanypeople.X]XXXb  ////  #XbXXX]#//Boersema//and//Zwaga//(1989)identifiedfivewarningmessagesthatwereusedtoreducethe   hazardousbehaviorofswimmingpoolslideusers.Foreachwarning,ninesymbolsweredeveloped.ThecomprehensibilityofthesymbolswastestedusinganevaluationprocedurebasedtoalargeextentontheInternationalStandardsOrganizationtestingproceduretodeterminethecomprehensibilityofpublicinformationsymbols(//ISO//9186,ProceduresfortheDevelopmentand   TestingofPublicInformationSymbols).Respondentswereswimmingpoolusersbetween7and19 p  yearsofage.Sevenacceptablesymbolswerefoundreferringtofourofthefivewarnings.////  CollinsandLerner(//1982b//)evaluatedU.S.participantsunderstandabilityoftwentyfiveinternationallyproposedsymbolsforfiresafetyalerting.Thepoorperformanceofsomecriticalsymbolssuchas exitwasnoted,andsomepotentiallydangerousconfusionsinmeaningwererevealed.Asaresult,theresearchersadviseincorporatingtestingproceduresasintegralpartsofthesymboldevelopmentandstandardizationprocessbeforewidespreadadoption.////  //Frantz//,Miller,and//Lehto//(1994)examinedtheimpactoftheflameandpoisonwarningsymbolsprescribedbytheCanadiangovernment.Theresultssuggestthatalthoughthegenericmeaningsofthesetwosymbolsarewellunderstood,peoplehavedifficultyinferringthespecificsafetyprecautionsmostnecessaryforaparticularproduct.Inthetestcaseofextremelyflammableflooradhesive,usersapparentlydidnotrealizethatadhesivevapors,ratherthantheadhesiveitself,posethefirehazard.X]XXXb  ////  #XbXXX]#Theauthorsofthe//FMC//CorporationProductSafetySignandLabelSystemmanualbelieve p thatcommunicationeffectivenessbetweenagreatercrosssectionofthepopulation,bothnationallyandinternationally,couldbeincreasedbyusingpictorialorsymboliclanguageinplaceof,orasasupplementto,writtenwords.ThepictorialsarecombinedwithwordsandcolorsinspecificallydesignedformatsintendedX]XXXbԀ#XbXXX]'#topresentcomprehensivehazardinformationinanorderlyand _!0 understandablemanner.X]XXXb Q"" ////  #XbXXX]#X]XXXbJayneandBoles(1993)investigatedwhetherdifferentwarningdesigns,specificallythose 1$  withsymbols,affectcompliancerates.Fiveconditionsweretested:averbalwarning,apictographswarningwithacircleenclosingeachgraphic,apictographswarningwithatriangleonitsvertexenclosingeachgraphic,awarningwithbothwordsandpictographs,andacontrol(nowarning).Participantsperformedachemistrylaboratorytaskusingasetofinstructionsthatcontainedoneofthefiveconditions.Thewarningsinstructedthemtowearsafetygoggles,maskandgloves.Allfourwarningconditionshadsignificantlygreatercompliancethanthenowarningcondition.Asignificantmaineffectwasfoundforthe"presenceofpictographs"variable,suggestingtotheauthorsthattheadditionofpictographstoaverbalwarningwillincreasecompliancerates.Theenclosureshapemadenodifferenceincompliancerates,despiteresearchthatindicatesthatunstableshapesarepreferredaswarningenclosures.Theauthorsechoed_the -b'* recommendationsofHorstthatconductingbehavioralstudieswaspreferabletorelyingonpreferencedata.#XbXXX]L#  //  // w$   w$}24O  3  .  1  .  3  .  2  3  0    Shapeoflabelsandwarnings e w$}݌(#(# Ќ    Thepropershapeofwarningsignshasbeenevaluated.Thetrianglearrangedonitsvertexwasthepreferredshape,withthecircledespiteitscommonusescoringamongtheleastpreferredwarningshapes(Riley,//_Cochran_//,and//_Ballard_//,//1982a//).Similarly,theConsumerProductsSafetyCommission(FederalRegister,May3,1996)testedseveral//_pictograms_//foreffectivenessinwarningaboutthehazardsofcarbonmonoxidepoisoningfromburningcharcoalbriquetsinconfinedspaces.TheyfollowedtheANSI//Z535.3//1991protocolfortestingandfoundthatthestandardslashrecommendedbyANSIwasnotaseffectiveasan Xforwarningconsumersofprohibiteduses.X]XXXb P  ////  WorkbyYoung(1993)focusedonincreasingthenoticeabilityofwarningsbymanipulatingfourvariables:pictorial,color,signaliconandborder.Subjectsviewed96simulatedalcohollabelsonacomputer,halfwithawarningandhalfwithout.Theresultsshowedthatwarningscontainingapictorial,color,oraniconhadsignificantlyfasterresponsetimesthanwarningswithoutthem.However,theadditionofaborderdidnotimproveresponsetimes.Moredetailedanalysesdemonstratedthatpictorials,colorandiconscanenhancethenoticeabilityofwarninginformationandinteractwitheachotherinsuchawaytothattheyshouldnotbeused_indiscriminantly_Ԁwithoutadequateknowledgeoftheinteractions.//#XbXXX]2#//  Warninglabelsaremostoftenrectangularinshape.Theorientationoftherectangularshapeisdeterminedbytheformatofthelabelandthenumberofelementscontainedwithinthelabel(i.e.,signalword,_pictograms_,messagetext).Theshapeofanysymbolscontainedwithinthelabel,however,aredrawnfromthepoolofshapescommonlyassociatedwithsafety,suchasoctagons,diamonds,triangles,andcircles.Whiletheshapesoftheelementswithinlabelsaresomewhatstandardizedandcommonlychosenforpromptuserrecognition,theeffectsoftheshapeofthelabelitselfhavenotbeenscrutinizedasclosely//_Kotwal_//andLerner,1995).Rectangularshapesarepredominantlyusedbecauseoftheirsimplicity.////  Besidesallowingforefficientuseofspace,theshapeofalabelmayenhancetheconspicuousnessofawarning,evenifthetargeteduserisnotactivelysearchingforsafetyinformation.Thismaybebecausethepublicmaybecomemoresensitivetocertainshapestimuliwhichtheyhavecometoassociatewithdangermuchastheydowithcolor.Asaresultofthisperceptuallearningphenomenon,//_Cunitz_//(1992)suggestedthatstandardizedshapes(e.g.,octagonshapedstopsign,triangleshapedyieldsign)cantakeongreaterattentiongettingcapabilitiesthanothershapes.////  //_ISO_//9186,ProceduresfortheDevelopmentandTestingofPublicInformationSymbols *$' specificallyrecommendstheuseofsurroundshapes.Thedefinedshapesareusedtodistinguishdifferentalertingcategories.However,theseshapesonlysurroundthe//_pictogram_//portionofthewarning.Ifaverbalmessageisincludedinthewarning,itisnotincludedinthesurroundshape. -`'* _ANSI//Z535.3//(1991)doesstatethattheperimeterofthelabelpanelsmaytakeonthesurroundshapeofsafetysymbols,althoughjustificationsforthisallowancearenotspecified.SafetyshapesrecommendedbyANSI//Z535.3//(1991)and//ISO//9186,alongwiththeirassociatedmeaningsareasfollows:////      S"0 `   Equilateraltrianglerestingonbase:Warningandhazardalerting Sn݌ ` (#` (# Ќ        -"0 `   Circle:Mandatoryaction -H݌ ` (#` (# Ќ  0      "0` (#(#  Circlewith45degreeslashfromupperlefttobottomright:Prohibitedaction ݌ ` (#` (# Ќ  0      "0` (#(#  Squareoroblongasnecessarytoaccommodatetext:Information ݌ ` (#` (# Ќ    Allotherrecommendationsincurrentguidelinesregardingshapesofelementsareactuallyreferringtoshapesofsymbolswhichareincorporatedaselementswithintherectangularbordersofthelabel(//Kotwal//andLerner,1995).X]XXXb o@     w$   w$24O  3  .  1  .  3  .  3  3  0 `   UseofMultipleMedia  w$݌O ` (#` (# Ќ    Therehavebeenseveralimportantstudiesthatpointtothevalueofpresentingmessagesinseveralmediasimultaneously.Forinstance,asymbolmaybecombinedwithwritteninformationorasignalwordtoreinforcetheeffect.Inonestudy,fiftyeightsubjectswereshownrandomlyorderedfacsimilesofstandarddangersignsrequiredbyOSHAandratedthesignson13dimensionsrelatedtoperceivedeffectiveness.Signscontainingahazardlabelandinstructions(e.g.,GASOLINENOSMOKING)wereratedasleastlikelytoberecalledatalatertime;however,theywereratedaseasiesttounderstand,mostinformative,andmostlikelytobecompliedwith.Signscontainingahazardlabelonly(e.g.,POISON)wereratedasleastinformativeandmostdifficulttounderstand;however,theywereratedasmostlikelytoberecalled,asdepictingahighdegreeofdanger,andlikelytobecompliedwith.Signscontaininginstructionsonly(DONOTENTER)wereratedasgenerallyleasteffective.////  #XbXXX]#Dingus,//Hunn//and//Wreggit//(1993)foundthattheinclusionofprotectivedevicesinthe o @ packagingofhazardousconsumerproducts(e.g.glovesinhaircoloringproducts)decreasedthecostofcomplyingwithwarningsspecifyingtheiruseandled,consequently,tomorecompliance.SeethediscussioninSection3.1.5.8,CostofCompliance.  ?#    [   [ 24O  3  .  1  .  4  3  0    READABILITY  [ 7 ݌%!(#(# Ќ    w$   w$0!24O  3  .  1  .  4  .  1  3  0    Legibility w$0![!݌& #(#(# Ќ  #!  Legibilityinvolvesboththetypographicalelementsandthesurfaceuponwhichthemessageisprinted(Hale,1991).ThiswasabundantlyillustratedwhenCongressrequiredahealthwarningontobaccopackagesbutprovidedlittleinthewayofspecifications.Thelabelsproducedcouldscarcelybereadinmanyinstances,duetotypesizeandthecontrastbetweentheinkandpaperchosen.In1988,Congresspassedasimilarrequirementforbeerandwinebutthistimeincludedspecificrequirements,includingthatthefirsttwowordsmustbe  GOVERNMENTWARNING , -`'* setinboldfacecapitals.  AstudycommissionedbythePublicHealthService(1990)lookedattheeffectivenessoftheseproscribedwarningsandfoundthatwhenthecolorofthetypographyandthecolorofthebackgroundweremeasured,aContrastRatio(CR)couldbederived,whichprovedtobeavalidmeasureoflegibility.SpecimenswithhighCRvalueswerelesslegible,despitecontainingthesamewarninginthesamesizetype.Hale(1991)concludedthatevenwithreasonableattemptstomandatelegibilitytherehavebeenanumberoffactorsnotspecifiedthatcanhavesignificanteffects. Ifregulatorsandotherswhowishtoformulaterulesforlegibilityaretoenjoyevenmodestsuccess,itisclearthattheywillneedtheassistanceofwelldefinedstandardscoveringalltheattributesdescribed(Hale,1991).////  //Howett//(1983)derivedaformulagivingtheletterstrokewidthneededforlegibilityofwordsonasignatanygivendistancebyanobserverwithanygivenvisualacuity.Thestrokewidth,inturn,determinesthelettersize,dependinguponthecharacteristicsofthetypefaceused.Thederivationisstrictlymathematicalandisbasedontheassumptionthatbeyondadistanceofafewmeters,aperson'svisualacuityisspecifiablebyafixedvisualangle,independentofthedistance.Theauthordoesprovidegraphsforcorrectingthecriticalstrokewidthfornonstandardcontrastorbackgroundluminancebasedonabodyofdataonvisualacuityasafunctionofcontrastandbackgroundluminance.   w$   w$+24O  3  .  1  .  4  .  2  3  0    Readinglevelevaluations +w$+@+݌(#(# Ќ    Dangerousaccidentsmayoccurwhenpeoplemisreadwrittenmaterials,consequentlymaterialsmustmatchthereadingleveloftheintendedaudience.Thereareseveralinternationallyrecognizedinstrumentsformeasuringthereadinglevelofwrittenmaterialsbutnoneofthemdealwiththecomprehensibilityofthematerials,asindicatedearlier.//Kreindler//and//Luchsinger//(1978)suggestdeterminingthereadinggradelevelofallsafetymaterialsusingstandardizedmethodssuchasthe//FORCAST//formulaorthe//Flesch//ReadingEaseIndex.//Laughery//and//Brelsford//(1996)repT~Dortedcommonlyfindingproductwarningswrittenatsignificantlyhighergradelevelthanthegrade4to6rangethatisusuallyrecommendedforgeneraltargetaudiences.AdiscussionofreadinglevelmeasuresandtheirapplicationtothedesignofinstructionsandwarningscanbefoundinDuffy(1985).    w$   w$~024O  3  .  1  .  4  .  3  3  0    TypeFontsandTypeSizew$~00݌&$"%(#(# Ќ  q0   BraunandSilver(1992)lookedatoverthecounterandprescriptiondrugsandexaminedcertainvariablesassociatedwithlegibility,namelyfonttype,fontweight,pointsizecontrastbetweenthesignalwordandthemainbodyofthewarning.TheresultsindicatedthatparticipantsweremorelikelytoreadthewarninginHelveticatype(asansserifface)thaninTimesor//Goudy//(bothseriffaces).Timeswasmorelikelytobereadthan//Goudy//.BoldtypewasmorelikelytobereadthanRomantype.Therewasagreaterlikelihoodofreadingthewarningwhenthemainbodywasin10pointsizeascomparedto8pointsize.A2pointsizedifferencebetweenthesignalwordandthe -)- mainbodyofthewarningproducedagreaterlikelihoodofreadingthewarningovera4pointsizedifference.Onepossibilityforthisresultisthatthe4pointsizedifferenceminimizestheimportanceofthemainbodyofthewarning,thereforemakingonlythesignalwordsalient.////  Lookingatpesticidelabeling,Silver,KlineandBraun(1994)obtaineddifferentresultsfromtheirearlierworkwithprescriptiondrugs.Thevariablesusedwerethesame:fonttype,pointsizecontrastbetweenthesignalwordandthemainbodyofthewarning(signalwordtextsizedifference),andpointsizeofthesignalword.Collegestudentsrated36insecticidelabelsthatcontainedawarningwhichvariedacrossalllevelsofthethreevariables.ResultsshowedthatCenturySchoolbookwasperceivedasmorereadablethan//Bookman//orHelvetica.Moreover,greaterperceivedreadabilitywasobtainedwhentherewasnodifferencebetweenthepointsizesofthesignalwordandthebodyofthewarning.Perceivedreadabilityandperceivedhazardousnessdecreasedasthesignalwordtextsizedifferenceincreased.Therewasgreaterperceivedreadabilityandperceivedhazardousnesswhenthesignalwordwasprintedin14pointtypethanin12or10pointtypes.   [   [924O  3  .  1  .  5  3  0    PLACEMENTOFINFORMATION 9[9 :݌/(#(# Ќ  X]XXXb  Oneofthekeyvariablesonwarningeffectivenessiswheretheinformationisplaced.Thisincludesplacementofinformationwithinalabelaswellasthephysicalplacementofawarningsigninproximitytothehazardthewarningdescribes.Intestsontheeffectivenessofthreedifferentwarninglabeldesignsforaportableelectricheater,researchersGill,//Barbera//,and//Precht//(1993#XbXXX]:#X]XXXb)foundthatthedesignrequiringtheusertointeractwiththelabelinordertousetheheater  wasmoreeffectiveinattractingtheusersattentionthatatraditionallabeloracolorcoded"skipass"labelattachednearthemaleendoftheelectriccord.However,noneofthemwereeffectiveinmediatingsafeuserbehavior.////  Anexperimentonthenoticeabilityofwarningsonalcoholicbeveragecontainersindicatedthatwarningsonthefrontlabelwerefoundmorequicklythanwarningsappearinginanyotherlocation.Also,warningsprintedhorizontallywerefoundmorequicklythanwarningsprintedvertically.Thesignalphrase("GovernmentWarning"),aswellastheamountof"noise"orclutteronthesurroundinglabel,significantlyinfluencedwarningdetectiontimes.Theauthorssuggestedthatpropermanipulationofthesedesignfeaturescouldmakethemandatedwarningmorenoticeable(Godfreyetal.1993)#XbXXX]<#X]XXXb.#XbXXX]A#X]XXXb %! //#XbXXX]WA#//  X]XXXbGodfreyandcolleagues(//1993c//)alsolookedattheeffectofthephysicalsizeandlocation & # ofthewarnings.Warningswerepostedonacopymachine,apublictelephone,awaterfountain,andtwosetsofdoors.Mostpeopledidnotusethecopymachineandthetelephonewhentheyhadthewarningsonthem.Thewarningonthewaterfountainwasnoteffectivewhenasingle,smallwarningwasused.Whenthatplusalarger,moreforcefulwarningwasused,mostpeopledidnotdrinkfromthewaterfountain.Thewarningsonthedoorswerenotobeyedunlesstherewasaconvenientalternativeexitavailable.Warningswithalowcostofcomplianceareapparentlymoreeffective.Additionally,thewarningscannotbetoosmallandmustbeplacedstrategically. -`'* ////  Aparadigmwasdevelopedtoexaminetheeffectivenessofwarningsinalaboratorytask(//Wogalter//,//Fontenelle//,and//Laughery//,1993).Compliance(useofmaskandgloves)wasaffectedbytheinclusionofthewarningaswellasbyitslocation.Greatestcomplianceoccurredwhenthewarningwasplacedpriortotheinstructions.Theadditionofaprintedstatementplacedbeforetheinstructions(withwarningattheend)toreadthroughtheinstructionsbeforebeginningproducedintermediatecompliancethatwasnotsignificantlydifferentfromthewarningbeginningandendconditions.Observationrevealedthatwhenthewarningmessagewasattheendoftheinstructionssubjectscompliedonlywhentheysawthewarningmessagebeforestartingthetask.Theseresultsindicatethatifwarningsareplacedinfrontofinstructionstheconsumerismorelikelytoreadandcomply.#XbXXX]A# `  // //  X]XXXbAnotherchemistrylabformatwasusedby//Wogalter//,//Kalsher//,and//Racicot//(1992)to o@  examinetheefficacyoftwowarningrelatedfactorstoproducecautionarybehavior.Experiment1comparedtheeffectsofapostedsignwarningandawithininstructionwarningonbehavioralcompliance.Theresultsshowedthatawarningembeddedinasetoftaskinstructionsproducedsignificantlygreatercompliance(thewearingofprotectivegear)thanasimilar,largerwarningpostedasasignnearby.Experiment2reexaminedtheeffectoflocationandalsoexaminedtheinfluenceofthepresenceversusabsenceofpictorials.TheresultsofExperiment2confirmedthelocationeffectofExperiment1.Noinfluenceofpictorialswasnoted,althoughtherewasa//nonsignificant//increaseincompliancewhenpictorialswereaddedtothewithininstructionwarning.Theresultsindicatedthatwarningplacementisimportantforelicitingbehavioralcompliancetosafetymessages.Explanationsuchasdifferenceinfieldofviewandperceivedrelevancearediscussed.#XbXXX]WI#   ////  //FMC//Corporation(1985),aspartoftheirsafetyprogram,emphasizedtheappropriate p placementofsafetysignsandlabelstoreducetheoccurrenceofaccidents.//Frantz//(1992)foundthatcontrarytocurrent,recommendedpractice,substantiallymoresubjectsreadandcompliedwithwarningsandinstructionsthatappearedinthedirectionsforuseratherthanthe precautionssection.Onaverage,movinganinstructionfromtheprecautionsintothedirectionsincreasedthereadingratefrom37%to89%andincreasedthecomplianceratefrom48%to83%.   [   [P24O  3  .  1  .  6  3  0    AUDIENCEVARIABLES P [P Q݌/$ (#(# Ќ    w$   w$ R24O  3  .  1  .  6  .  1  3  0    Generalconsiderations Q w$ R5R݌&"(#(# Ќ    McCarthyandcolleagues(1984)examinedapproximately400articlesontheeffectivenessofonproductwarninglabelsandwereabletocategorizetheliteraturebasedonthemannerinwhichthewarningissuewasaddressed:0  Qualitativeeditorial.Theseappeartobebasedontheexpectationthatprovidingsafety +%( informationisrelatedtoanincreaseinsafetybehavior. (#(# Ѐ0  Recognitionrecall.Thesearethemostrepresentedquantitativeworksintheliterature -`'* addressingrecognitionofthewarningmessageandabilitytorecallthatmessageatalatertime. (#(# 0  Standardsandlabelsystems.Articlesonconsensus,governmental,andindustrialstandard  systems. (#(# 0  //Quantification//ofeffectiveness.Theseareamongtheleastrepresented. (#(#     Theauthorsconcludedfor//nonquantitative//studiesthatnotoneofthedesignvariablesdiscussedinlabelcreation isrelatedinanyknownwaytotheactualeffectivenessofwarningsinchangingbehaviorandreducinginjuries.Further,theyfeltthatallavailablequantitativestudiesshowedthewarningsunderinvestigationtohavenoimpactonsafety.Horst(1993)reviewedtheliteratureofexperimentalpsychologyandcommunicationtheoryandpointedoutthatpeoplelearntofilteroutmostoftheinformationthatbombardsoursensesandtoattendtothingsthattheyhavelearnedtoviewasimportantorinteresting.Ifsafetyinformationisperceivedasfamiliarorredundant,itmaysimplybedroppedfromshorttermmemory,andhavenofurthereffectonbehavior.Alternatively,iftheinformationisperceivedasinconsistentwithexistingknowledgeandbeliefs,itmayberejectedasnotcredible.Eveniftheinformationispreviouslyunknownandisacceptedastrue,therecipientmaybeinsufficientlymotivatedtoalterbehavior.////  Similarly,Millerand//Lehto//(1986)lookedat388referencesourcesonthesubjectofwarningsandfoundonly10sourcesthattheyconsideredtrue//analyses//ofwarningseffectiveness,andofthese,only6actuallyciteanyexperimentalresults.//Purswell//,//Krenek//,and//Dorris//(1993)donotfindthispaucitysurprisingbecauseofthedifficultiesassociatedwithperformingthesestudies.Thereisthegeneralproblemofcollectingexperimentalobservationsandthespecificproblemofinadvertentlyintroducinganexperimentalbiasbecausethesubjectisinfluencedtoreadthewarningandbehavesafelyifheorsheperceivesthatthepurposeofthestudyistoevaluatewarningeffectiveness.////  Horst(1993)alsoreportedthattherearewidelyencounteredguidelinesbasedprimarilyonpreferencestudiesthatwarningsshouldcomprisefourelements:astandardsignalword,statementsofthehazard,thepotentialconsequences,adhowtoavoidthehazard.Horstetal.claimitiserroneoustosuggestthatallofthisinformationmustbeprovidedtoaffectbehaviororthatinclusionofallofthisinformationwillguaranteeeffectiveness.Theauthorsgavetheexamplesofstopsignsandrestroomsignsneitherofwhichcarryallfourelementsbutareeffectivewarnings.Stopsignsareeffectivebecausethenegativeconsequencesoffailingtorespondappropriatelyareuniversallyfamiliarandhighlyprobable.Similarly,thehighprobabilityofsocialembarrassmentisclearlyamotivatorintheeffectivenessofrestroomsigns.   w$   w$@b24O  3  .  1  .  6  .  2  3  0    Gender 3b w$@bkb݌("%(#(# Ќ    //Laughery//and//Brelsford//(1993)reviewedtheliteratureandreported reasonablycleartrendsthatindicatewomenreportagreaterlikelihoodthanmentolookforandreadwarnings.Theyarealsomorelikelytocomplywithwarnings.//  -`'* ////   w$   w$d24O  3  .  1  .  6  .  3  3  0    Age d w$de݌(#(# Ќ    //Kotwal//andLerner(1995)concludedfromtheircomprehensiveandexcellentreviewofliteratureonwarninglabelsthattheageoftheuserappearstohaveaneffectontheperceivedlevelofhazardconveyedbycommonsignalwords.Leonard,Hill,and//Karnes//(1989)foundsignificanteffectsbetweenyoungercollegestudentsandoldercollegestudents(overage25).//Kotwal//and   Lerner(1995)summarizedtheirfindingswiththefollowingstatement:0  Oldersubjectsgenerallyusedsignalwordsthatimpliedgreaterhazardstorepresenttheamountofriskinvolvedinagivensituation.Sinceolderusershaveindicatedthatagivensignalwordimpliesalowerlevelofhazardthanthesamewordimpliesforyoungerusers,itmaybenecessarytoapplyrelativelystrongsignalwordsforolderusersinordertoconnoteagivenlevelofhazard. (#(#   //Desaulniers//(1991)reportedthatolderpeople,40andabove,aremorelikelytotakeprecautionsinresponsetowarningsasreflectedintheirbehavioralintentions.However,Collins / &Lerner,(1982)foundthatoldersubjectsexhibitedlowerlevelsofcomprehensionforsafetysignsinvolvingpictorials.For//Laughery//and//Brelsford//(1993)thesestudiestakentogethersuggestthatolderreceiversaremoreinfluencedbywarnings,butgreaterattentiontoissuesofcomprehensionmaybenecessary. !XbXXXb   w$   w$^l24O  3  .  1  .  6  .  4  3  0    Hazardandriskperceptionw$^ll݌(#(# Ќ  Ql   Aconsistentfindinginwarningresearchisthatpeoplesperceptionofthehazardousnessassociatedwithaproductorsituationcandeterminetheeffectivenessofawarningorlabel(//Laughery//and//Brelsford//,1993).#XbXX!Xb2l#X]XXXbManyresearchershaveacknowledgedthattheactualdesignof ` thewarninglabelmaynotbeasimportantastheexpectationsthattheuserbringstothesituation(//Kotwal//andLerner,1995).#XbXXX]n#ԀResearchhasindicatedthatperceivedseverityofconsequenceshasa q B stronginfluenceonbehavioralintentions.Numerousstudieshaveshownthatthegreatertheaprioriperceptionofhazardousness,themorelikelypeoplewilllookforandreadawarningandthemorelikelytheywillcomplybytakingprecautions(Donnerand//Brelsford//,1988;Friedman,1988;//Otsubo//,1988).Additionally,themoreexplicitthewarningisabouttheseverityofinjuryconsequences,thegreatertheperceivedhazardousnessandrecallofwarninginformation.//X]XXXb//  //Purswell//,//Schlegel//,//Kejriwal//,and//Sashi//#XbXXX]\r#Ԁ(1993)reportedthattherehavebeensurprisingly ' # fewattemptstodevelopausefulmeasureofrisktakingbehavior,giventheimportanceofthisfactorinsafebehavior.TheyX]XXXbԀdevelopedamodeltopredictwhetheraconsumerwoulduseaproduct ("% safelyasafunctionofsixteendifferentindividualvariable.Subjectswerepresentedwithfourconsumerproductstouseinanexperimentalsettingwherethetruepurposeofthestudywasconcealed.//Discriminant//analysiswasusedtodevelopapredictionmodeltoclassifysubjectsintocategoriesofsafeorunsafebehavior.Predictionaccuracyrangedfrom6886percentfordifferenttypesof_behavior. -h'* ////  Otherresearcherslookedatpeople'swillingnesstoreadwarningsonhouseholdpestcontrolproductsandtheirlikelihoodofpurchasingtheseproducts(Silverand//_Wogalter_//,//1991b//).Seventycollegeagedand20olderparticipantsexaminedasetofpestcontrolproductsandrespondedtoaquestionnaireassessingperceptionsoftheproducts,thepackaging,andthewarnings.Resultsshowthatproducthazardousness,warningunderstandability,andwarningattractivenesswerestronglyrelatedtopeople'swillingnesstoreadthewarnings.Adifferentsetofvariableswasrelatedtopurchasingintentions.Participantsreportedgreaterwillingnesstopurchaseproductsthatweremorefamiliarandhadmoreattractivepackaging.Participantsweremorewillingtoreadwarningsthatcontainedmorestatementsandhadreadabilityscoresathighergradelevels,aresultthatappearedtobeduetotheircommonrelationshipwithperceivedhazard.Thefindingssuggestedthatmanufacturerscanplaceappropriateandeffectivewarningsonpestcontrolproductswithoutnecessarilyreducingbuyingintentions.#XbXXX]s#X]XXXb P  ////  #XbXXX]+{#//_Bogett_//andRodriguez(1987)exploredtheinfluenceofaperceptionofdangerparticularly _0  regardingproductbornewarningsandsafetyinstructionalprograms.Theresultsoftheirstudy,togetherwithsupportingdatafromtheliteratureimpliedthataperceptionofdanger(i.e.,anunacceptableriskoflossorinjury)mustexistinordertoelevateapersonssafetybehavior. FollowingPages:Tablefrom//_Kotwal_//andLerner(1995)SummaryofResearchFindings  RegardingUserAge,Gender,PerceivedHazardousness,andFamiliarityWithProduct      ` XXb!*+, d dzdd zdd ~dd ~dd dd dd dd dd dd Zdd Z dd c"d(#(#, d ,ltd , td ,- t +  3---#  h3@tt AUTHOR 8-"-##Q"---# 8@ { TITLE 8-"-##Q"-"-# 8@#METHODOLOGY 8-"--##Q"-"-# 8@:FINDINGS/@<<9CONCLUSIONS J-;#Q" 0 h -"--# xJGodfrey,S.S.,//_Allender_//,L.,_Laughery_,_K.R._,andSmith,//_V.L._//(1983) 2"#I"%- 2Warningmessages:Willtheconsumerbothertolook?Proceedingsof * theHumanFactorsSociety27thAnnualMeeting,950 #. 954. 2"##/" 2EXP1̄Subjects:32undergraduatecollegestudents̄Subjectswereaskedtoimaginethemselvesinvariouspurchasingsituationsandwerepresented8genericcommonconsumerproducts(plantfood,ovencleaner,pesticides)inabookletformat̄Subjectswereaskedtorateproductsonfourhazardscales(inhalation,swallowing,skinirritation,andoverall),threelikelihoodscales(likelihoodtolookforwarningmessage,toreadthemessage,andtocomplywiththemessage),andonfamiliaritywiththeproduct 6"-#*$B" 6ЄFoundthatsubjectsweremorelikelytolookforwarningsonlessfamiliarandmorehazardousproducts̄Foundthatfemalesweremorelikelytolookforawarningthanmales̄However,forproductsthatweremosthazardous(e.g.,pesticides),subjects'degreeoffamiliaritydidnotmatter;subjectsstillreportedthattheywouldlookforandreadwarningsregardlessoftheirdegreeoffamiliaritywiththeproduct J-;#q%BN x "- XJ 2"## - 2 2""## " 2EXP2̄Subjects:88undergraduatecollegestudents ℀Studywasconductedinamannersimilartoalabpractical,using12different"stations"̄40householdproductswereusedintheexperiment;eachsubjectsaw12ofthe40products̄Subjectswereaskedtorespondtoquestionsregardingfamiliaritywiththeproduct,similaritytootherproducts,frequencyofuse,recencyofuse,andperceivedhazardousness 6""-#/ "" 6ЄFindingsindicatethatsubjectshavelowerhazardperceptionsforproductsthatarefamiliar J"-;#  X ""- :JGodfrey,S.S.and//_Laughery_//,//_K.R._//(1984) 2""#"- 2Thebiasingeffectsofproductfamiliarityonconsumers'awarenessofhazard.Proceedingsof l! theHumanFactorsSociety28thAnnualMeeting,483 /% 486. 2""#^&"" 2ЄSubjects:110femalestudents(mostwereundergraduateandgraduatecollegestudents;17werehighschoolstudents)̄Surveyedwomenontheirawarenessofthehazardsoftamponuse,knowledgeofthesymptomsoftoxicshocksyndrome,andawarenessofwarnings 6""-#0"" 6ЄFoundthatwomenwhoweremorefamiliarwithtamponproductswerelesslikelytonoticewarningswhentheyswitchedbrands̄Datasuggestthatbeingfamiliarwiththeproductorasimilarproductleadstolowerperceptionofhazard L"-";#=7 : ""- P LLeonard,S.D.,Matthews,D.,and//_Karnes_//,//_E.W._//(1986) 4"""# D"-" 4Howdoesthepopulationinterpretwarningsignals?Proceedingsof K   theHumanFactorsSociety30thAnnualMeeting,116 s 120. 4"""#=""" 4Subjects:368collegestudents̄Studiedwhetherdifferentsignalwordsproducedifferentperceptionsofrisk̄Subjectsviewedwarningsignswhichvariedaccordingtosignalword(DANGER,WARNING,CAUTION),signalwordcolor(REDorBLACK),signalwordsize,andinformationpertainingtotheconsequencesofdisregardingthewarning(consequencevs.noconsequence)̄Subjectsweretoldwherethesignsmightbelocatedandwereprovidedscenariosinwhichthewarningsignmighthave relevance;subjectsratedtheperceivedriskassociatedwitheachsignandthelikelihoodthattheywouldobeythesigndisplayedinthescenariosprovided 6"""-#I)""" 6ЄFoundthat,givenadescriptionofthehazardbeingwarnedagainst,subjectsdidnotrateriskdifferentiallyasafunctionofthesignalwords̄Neitherthesizeofthesignalwordnorthecolorsignificantlyaffectedtheperceptionofrisk̄Theperceptionofriskseemedtobepredicatedontheinformationcontentofthesign̄Concludedthatinformationaboutconsequencescouldraisetheaverageperceptionofrisk̄Notethatonepossibleexplanationoftheresultsisthatpriorexperiencehasgeneratedacertainlevelofperceptionofrisk,basedonpersonalexperiencesandpossiblyonvicariousexperiencesthroughthecommunicationsmedia L-";#W<  P """- xLLeonard,D.C.,//_Ponsi_//,//_K.A._//,Silver,N.C.,and//_Wogalter_//,M.S.(1989) 4""#"PE-" 4Pestcontrolproducts:Readingwarningsandpurchasingintentions,Proceedingsof W!L theHumanFactorsSociety33rdAnnualMeeting,436 $P 440. 4""#I%Q"" 4ЄSubjects:70undergraduatecollegestudents(1719yrs)and20olderadultundergraduatestudents(meanage37yrs,stddev7.7yrs)̄Warningson22pestcontrolproductswereevaluated̄Subjectsweregivenaquestionnairetoassessperceptionsoftheproduct'spackaging,labeling,andwarnings;responseswererecordedusinga9point//_Likert_//scale(0to8)̄Examinedwhetherseveralobjectivemeasuresofthewarningreadability(statements,words,gradelevel)wouldberelatedtothewillingnesstoreadvariable 6""-#*4"d"" 6ЄWarningswithmoreinformationandhavingmoredifficultmaterialmaycauseperceptionsofhazardousness̄Foundthatperceivedhazardousnessisanimportantdeterminantofwillingnesstoreadwarnings̄Hazardperceptionismoreimportantthanperceptionsoffamiliaritywithregardtowillingnesstoreadwarningsondangerousproducts J"-;##n x ""- J//_Wogalter_//,M.S.,//_Desaulniers_// ,//_D.R._//,and//_Brelsford_//,//_J.W._//,Jr.(1986) 2""# D"- 2Consumerproducts:Howarethehazardsperceived?Proceedingsof K   theHumanFactorsSociety31stAnnualMeeting,615 s 619. 2""#="" 2EXP1̄Subjects:28undergraduatecollegestudents̄Subjectswerepresented72genericproductsinalistformat̄Subjectsratedeachproductonsuchattributesasperceivedhazardousness,likelihoodofinjury,frequencyofuse,familiarity,levelofknowledgeofhazardsusinga9point//_Likert_//scale 6""-#6 "" 6ЄResultsindicatethatseverityofinjuryisthebestsinglepredictorofhazardperception̄Findingssuggestthatproductsthatarelessfrequentlyusedandlessfamiliarareperceivedtobemorehazardous H"--7#%  ""- 0H//_Wogalter_//,M.S.,//_Desaulniers_//,//_D.R._//,and//_Brelsford_//,//_J.W._//,Jr.(1986) 4""-#D,"-- 4Consumerproducts:Howarethehazardsperceived?Proceedingsof K 1 theHumanFactorsSociety31stAnnualMeeting,615 s5 619. 4""-#=6""- 4EXP2̄Subjects:70undergraduatecollegestudents̄Subjectswerepresentedalistof18productsandwereaskedtorateperceivedhazardousness,toindicatepossibleaccidentscenariosassociatedwitheachproductincludingtypeofinjury,severityofinjury,andlikelihoodofinjury 6""--#lA""- 6ЄResultsindicatedthatperceivedhazardousnesswasthemostimportantdeterminantofwillingnesstoreadwarnings̄Familiaritywiththeproductreducedthelikelihoodofreadingthewarning,butfamiliaritymaynotbeasimportantasperceivedhazardousness3)'I 0 0  ""-- 3#!~##XbX~#X]XXXb   w$   w$B24O  3  .  1  .  6  .  5  3  0 `   Stress 5 w$Bm݌VJ` (#` (# Ќ    Alongwiththeperceptionofhazard,anotherimportantexternalfactorwithrespecttowarningeffectivenessisstress.//_Magurno_//and//_Wogalter_//(1994)usedachemistrytaskformatinwhichthestressorwasacombinationoftwokindsofstress:timepressureandsocialevaluationbyanotherperson.Therewerefourconditionsevaluated:stress(lowerversushigher)andwarninglocation(postedsignvs.withintaskinstructions).Theresultsshowedthathigherstressproducedsignificantlylowercompliance.Warningplacementwasalsofoundtoproduceamaineffectonbehavioralcompliance.Moreparticipantscompliedwiththewithininstructionwarningthanthepostedsignwarning,eventhoughthesignwasover40timeslargerintermsofarea. #XbXXX]#   w$   w$24O  3  .  1  .  6  .  6  3  0    Productfamiliarity w$-݌A'U(#(# Ќ  X]XXXb  Astudyofonehundredstudentsusinghammersinalaboratorysettingresultedinnone !) W evennoticingthewarninglabelsonthehammers(//_Dorris_//and//_Purswell_//,1977).Theauthorsconcludedthatawarninglabelonafamiliarproductisnotnecessarilyasalientitemandmaybefilteredoutbytheusersattentionalmechanismsandneverprocessedataconsciouslevel.Godfreyand//_Laughery_//(//1993c//)foundthatwomenwerenotsolikelytonoticewarningsabouttoxicshocksyndromeonorintamponpackageswhentheyswitchedproducts.Theauthorsfound -6%\ arealbiasingeffectfromproductfamiliarity.#XbXXX]5#Ԁ / ////  //_DeJoy_//(1989)reviewedtheliteratureonwarningsandfoundsevenstudieswherefamiliaritydecreasedthelikelihoodofnoticing,reading,orcomplyingwithwarnings._Dejoy_Ԁnoted,however,thatthesefactorsdonotappeartobeasimportantastheusersproductrelatedexpectations.Similarly,interviewresearchintotheeffectivenessofwarningsignsonscaffoldingshowedthatbothinexperiencedandexperiencedsubjectswouldreacttoanewwarningonscaffoldingtheywereunfamiliarwithbutnottothesamenewwarningonequipmentthatwasfamiliartothem(Johnson,1993).Thisstudy,however,wasbasedonmeasuringbehavioralintentionthroughinterviews,notbehaviors.////  Manyotherstudieshavelookedattheeffectsoffamiliarityonwhetherornotsubjectswilllookfor,notice,andreadawarning.Godfrey,etal.(1989),Godfrey&_Laughery_Ԁ(1993c),_LaRue_&Cohen(1987),and_Otsubo_Ԁ(1988)haveallreportedthatpeoplearelesslikelytolookfor,noticeandreadwarningsonconsumerproductswithwhichtheyarefamiliarthantheyarewithunfamiliarproducts._Laughery_Ԁand_Brelsford_Ԁ(1993)consideredthefamiliarityeffectconsistentandrobustandsuggestedapossiblereason:themorepeopleuseaproductwithoutexperiencingasafetyproblem,thelesshazardoustheyperceivetheproducttobe.Intheauthorswords: productsthatareusedrepetitivelyposespecialwarningproblems.////  //_Ayres_//,etal.(1994)suggested,afteranextensiveliteraturereview,thatwarningsareunlikelytobeeffectiveunlessaseriesofconditionsaremet.Thefailureofmanyintendedwarnings,includingmostonproductwarninglabels,toreduceaccidentsreflectsthedifficultyofovercomingproblemsinherentintheiruse.   w$   w$N24O  3  .  1  .  6  .  7  3  0    Color A w$Ny݌r(#(# Ќ  X]XXXb  Coloraffectsbehavior.Braun,//_Sansing_//,andSilver(1994)pointedtoanextensiveliteratureontheeffectsofcolorbutnotethatvastmajorityofproductwarningresearchhasnotconsideredcolorbeyondtheapplicationofstandardsandguidelines.//_Dunlap_//,//_Granda_//,and//_Kustas_//(1986)surveyed1169subjectsacrossseveraldifferentlanguagegroupsincludingEnglish,German,Scandinavian,andSpanish.Subjectsratethecolorwordsred,orange,yellow,blue,green,and A# whiteaccordingtothelevelofperceivedhazard.Theresultsdemonstratedthatthehazardinformationcommunicatedbydifferentcolorsfollowedaconsistentpatternacrosslanguagegroups.Redresultedinthehighesthazardratingsfollowedbyorange,yellow,blue,green,andwhite,respectively.////  Researchershaveexaminedavarietyofattributesthatinfluenceawarning'sabilitytocommunicateimportantproducthazardsbut#XbXXX]r#Kline,Braun,Peterson,andSilver(1993)X]XXXbԀconsiderthe )#& attributeofcolortohavebeenneglectedintheliteratureonwarnings.Theauthorslookedattheappropriatenessofachromaticstimuliinproductlabelsandfoundthatcolorlabelswereperceivedasmorereadableandhazardousthanachromaticlabels.#XbXXX]ㄿ#Ԁ ,v&)  -h'*   ?+ 4 <DL!X?BraunandSilver(1994)examinedtheeffectofcoloroncompliancewithprintedwarnings.Theparticipantsinthestudyindicatedahigherlikelihoodofinjuryassociatedwithproductsprintedinredthangreenorblack.//_Barbera_//andGill(1987)attemptedtoassesstheabilityoftwowarninglabeldesignsonaconsumerproductto:(1)attracttheusersattention,and(2)altertheusersbehavior.Foronecondition,thewarninglabelconsistedofaonebythreeinchblackandwhitetagaffixedtothebackofaportableelectricheater.Thealternativeconditioninvolvedthesameproductbutwithathreebyfiveinchredandblackcolorcodedwarninglabelwithpictographs.Thissecondwarninglabelwasmountedattheendoftheheaterselectricalcordintheformofa skipasstag.Resultsrevealedneitherwarninglabeltobeeffective.Eventhe properwarninglabelcouldnotinducethesubjectstoreaditnoraltertheirbehavior.////  Buildingontheirpreviouswork,Gill,//_Barbera_//,and//_Precht_//(1993)movedawayfromsolelypassivelabeldesigns.Threedifferentwarninglabeldesignsforaportableelectricheaterweretested:(1)atraditionalnonhumanfactoredlabel;(2)acolorcoded skipasslabelattachednearthemaleendoftheelectriccord;and(3)acolorcoded interactivelabelthatrequiredtheusertointeractwiththelabelinordertousetheheater.Theresultsshowedthattheinteractivedesignwasmosteffectiveinattractingtheusersattention,butnoneofthemwereeffectiveinmediatingsafeuserbehavior,theultimategoal.////  Braun,//_Sansing_//,andSilver(1994b)examinedtheinteractionofsignalwordsandcolors.Asampleof30undergraduatesratedtheperceivedhazardof105signalword/colorcombinationsprintedinspecifichazardcolors.Ofthecolorsused,redconveyedthehighestlevelofperceivedhazardfollowedbyorange,black,green,andblue.Moreimportantly,however,itwasnotedthatasignalwordsuchasDEADLY_connotated_Ԁlesshazardwhenprintedingreenthanredink.////  Additionalstudiesintheareaofcolorandwarningssupporttheearlierfindingsthatthelevelofhazardcommunicatedbysignalwordsvariesasafunctionofthecolorinwhichtheyareprinted(Braun,Kennedy,andSilver,1994c).Importantly,thesefindingssuggestthatsignalwordandcolorcombinationscreateacontinuumofperceivedhazard.Thismodelofacontinuummayprovehelpfultodesignersofwarnings.   w$   w$24O  3  .  1  .  6  .  8  3  0    CostofCompliance w$݌?#(#(# Ќ    Costofcompliancereferstotheamountofeffortausermustexertinordertocomplywithawarning(//_Kotwal_//andLerner,1995).Thecosttotheuserofcomplyingwithawarningcanbereduced,forexample,byincludingprotectivewear(e.g.,gloves,mask,earplugs,etc.)withthehazardousproduct.//_Kotwal_//andLernerpointoutthattheissueofcostofcomplianceasaformalresearchtopicisrelativelynew,asmostoftheliteraturewhichaddressesthisissuehasbeenpublishedwithinthepastsixyears.Therefore,itisnotsurprisingthat,althoughseveralrecentarticleshavedemonstratedhowstronglycostcorrelateswithcompliance,costofcomplianceisnotevenmentionedinmajorindustrystandards(ANSI,1991;WestinghousePrintingDivision,1981;//_ISO_//,1984).//_Kotwal_//andLernerconcludethattheamountofeffortrequiredbytheusertocomplywiththewarninghasnotbeenafactorduringtheprocessofdesigningawarning,butrather,an -`'* _afterthoughttothedesign.However,thechangesincompliancerateswhichresultfromlowercosttotheuserprovidestrongevidencefortheinclusionofcostofcomplianceasadesignfactorinnewlabelingguidelines.////  Intuitively,thelowertheeffortrequiredbytheusertocomplywiththewarning,themoreimpressivetheincreaseincompliancerates.Inafieldstudy(Dingusetal.,1991)universitystudentsagreedtoevaluatethemarketingpotentialofa"new"cleaningproduct.Complianceinvolvedwearingrubbergloveswhileusingtheproduct.Subjectsinthelowcostconditionhadglovesprovidedwiththeproduct,whilethoseinthehighcostconditionhadnoglovesprovided.Theresultsshoweda25%compliancerateforthehighcostcondition,increasingto8788%complianceforthelowcostcondition.//Hunn//etal.(1992)foundthat"inadditiontoreducingcostofcompliance,theinclusionofglovesaspartofproductpackagingincreasedtheperceptionofdangerassociatedwiththeproductuse.Therefore,increasedcompliance,insomecases,isprobablyaresultoflowercost,aswellasanincreasedperceptionofdanger.////  Similarincreasesincompliancerateswerereportedby//Kotwal//andLernerduringtheirliteraturereview(1995).Increasesincomplianceratesrangeanywherefrom23%toalmost94%,dependingonthegivenexperimentalsituation.//Wogalter//,McKenna,andAllison,(1988)usedachemistrylaboratorysettingtodeterminehighandlowcostcompliancerates(17%and73%,respectively)forwearingprotectiveglovesandmasks.Studiesonaracquetballcourtfoundhighcostsresultedinnocompliancewithwearingprotective//eyewear//,whilelowcostgave60%compliance(HathawayandDingus,1992;Dingus,Hathaway,and//Hunn//,1991).////  Relatingthecostofcompliancetootherwarningfeatures,//Hunn//etal.(1992)foundthatlowcostofcompliancehadagreatereffectoncompliancethaneitherwarningcontentorwarninginteractivity.Additionally,Hathawayetal.(1992)foundthatthebenefitsofalowcostofcompliancecouldbeincreasedbyaddingspecificconsequenceinformationtothewarning.Theresearchersconcludedthatprovidingtheuserwithinformationregardingfrequencyandseverityofinjuriesassociatedwiththehazard(inthiscase,eyeinjurieswhileplayingracquetball),aswellasprovidingthetoolsrequiredtoexhibitsafebehavior(e.g.,protective//eyewear//)canmarkedlyimprovewarningeffectiveness.//X]XXXb Tablefrom//Kotwal//andLerner(1995) /$  SummaryofResearchFindingsRegardingCostofCompliance #XbXXX]_#!XbXXXb!X!Xb  %! *$"%Vdd d ltd l td - t - +,(#(#,( dd ,( dd ,( dd +  3---# / h3@@pp AUTHOR A-"-#,!r"---# A@@(TITLE A-"-#,!r"-"-# A@@;METHODOLOGY S-D,r" 0 h -"-# xSDingus,T.,Hathaway,J.,and//Hunn//,B.(1991) ;",! - ;Amostcriticalwarningvariable:Twodemonstrationsofthepowerfuleffectsofcostonwarningcompliance.Proceedingsofthe q B  HumanFactorsSociety35thAnnualMeeting,10341038. 4""# " 4EXP1̄Subjects:920racquetballplayersatlargecentersin2universities̄Manipulatedcostbyvaryingtheamountofeffortrequiredtoobtaineyewear̄Manipulatedthelevelofwarning ℀Measuredcomplianceasafunctionofwhetherornotsubjectsworeprotectiveeyewear̄Lowcostcondition=//eyewear//providedjustoutsidèdoor̀Middlecostcondition=subjectswererequiredtòwalk60feettocheckoutboothtoget//eyewear//̀Highcostcondition=no//eyewear//providedonsite J-;#G! x "" J ;",!z"- ; 2""#z#" 2EXP2̄Subjects:318universitystudents̄Experimentwaspresentedundertheguiseofamarketingstudy;subjectsthoughttheywereevaluatingthemarketingpotentialofa"new"cleaningproduct̄Measuredcomplianceasafunctionofwhetherornotsubjectsworerubbergloves̄Lowcostcondition=glovesprovided̀Highcostcondition=glovesnotprovided J"-;#"1 "" `JDingus,//T.A.//,//Wreggit//,S.S.,andHathaway,//J.A.//(1993)    R  l ;"",! "- ;Warningvariablesaffectingpersonalprotectiveequipmentuse.SafetyScience,16,655673. 2""# "" 2EXP2̄Subjects:224adultvolunteers̄Subjectsweregivenapackagecontainingaspraybottlewithcleaningsolution̄Manipulatedthreelevelsofinteractivity(nophysicalinteraction,"billboard",and"triggerguard"see//Hunn//andDingus,1992),labelcontent(genericproductinstructions,ANSIwarning,andANSIplusspecificconsequencewarning),andcostof compliance(lowcost:glovesandmaskprovided,highcost:noglovesandmaskprovided)̄Subjectsweregivenaquestionnaireafterusingtheproductforoneweek;wereaskedaboutproductsatisfaction,perceptionofrisk,andbehaviorindealingwiththeproduct̄Subjectswerealsoaskedwhattheythoughtthepurposeofthestudywas;ifthesubjectansweredthisquestioncorrectly,thatsubject'sdatawasincludedintheanalysis J"-;#q" ` "" JHathaway,J.andDingus,T.(1992) ;"",!# "- ;Theeffectsofcompliancecostandspecificconsequenceinformationontheuseofsafetyequipment.AccidentAnalysis&Prevention,   24(6),577584. 2""#K "" 2ЄSubjects:420racquetballplayersfromalargeuniversity(339males,81females);playerexperiencerangedfrom1sessionto13years;eachsubjectwasusedinonlyonecondition̄2x3completefactorialbetweensubjectsdesign̄Manipulatedcost̀Lowcost=eyeprotectionprovided̀Highcost=eyeprotectionnotprovided̄ManipulatedwarninginformatioǹBaseline=warningnotprovided̀ANSIstandardwarning̀ANSIstandardwarningplusspecificconsequencèinformation ℀Baselinecondition:strictlyobservation(i.e.,eyeprotectionnotprovided,nowarnings)̄Experimentalconditions:allmanipulationsofthetwoindependentvariables(i.e.,costofcomplianceandwarninginformation)3)' !   "" 3#!0#  XbX [   [24O  3  .  1  .  7  3  0    MEASUREMENTANDPROTOCOLS [݌ j#(#(# Ќ    Overtimethefocusofresearchonwarningshasshiftedfromadebateonwhetherwarningsworktosystematicinvestigationofthefactorsthatdoorcouldinfluencethebehaviorofusersofsafetyproduct.//Ayers//andhiscolleagues(1992)pointedoutthatthelogicaltestofawarningmustbereductionofthefrequencyand/orseverityofaccidentsandinjuries.Thisshouldbethekeymeasurement.////  Delegatesfromeightsocietiesincluding:theAmericanSocietyforTestingandMaterials,AmericanAssociationofTextileChemistsandColorists,American//Pharmaceuctical//Association,AmericanPsychologicalAssociation,IlluminatingEngineeringSociety,OpticalSocietyofAmerica,TextileColorCardAssociationoftheUnitedStates,andU.S.PharmacopoeiaConventioncollectivelyassembledintoagroupknownastheInterSocietyColorCouncil,or//ISCC//(//Billmeyer//,1991).The//ISCC//originallyassembledtostandardizecolortermsandcolornamesusedtodescribethecolorsofpharmaceuticals.Evenafterthecolorstandardizationproblemwassolved,theISCC %-$1 continuedtoexistwiththeintenttoadvancetheknowledgeanduseofcolorinart,science,andindustry.Thisorganizationmaybeabletoprovideimportantassistancewiththeinternationalharmonizationefforts.////  TheCommissionInternationaledelEclairage(CIE)sanctionstheinternationallyacceptedmethodofcolorspecificationandmeasurementthatisdesignedtomimichumancolorperception(Hale,1991).ANSIZ535isasetofstandardstoguidetechnicalcommunicatorsinthedevelopmentofeffectivehazardalertmessages.ANSIZ535.31991istheANSIstandardentitled CriteriaforSafetySymbolswhichcontainsatestmethodfordeterminingtheeffectivenessofapictogramandacriterionforsuccessof85%correctresponseswithnomorethan5%criticalconfusion.Thelattertermreferstowhenthemessageconveyedistheoppositeoftheintendedmessage,whichforsafetywarningscouldresultindeath.AscorebelowthisANSIleveldoesnotmeanthepictogramcannotbeusedbutthatitcannotbeusedaloneandmustbecoupledwithsomeothermedium,suchasawrittenwarning.ANSIZ535.4istheANSIstandardforProductSafetySignsandLabels.////  TheInternationalStandardsOrganizationissuedastandard,ISO9186,Proceduresforthe ? DevelopmentandTestingofPublicInformationSymbols,thatrecommendstestingmethodologies / toevaluatesymbolsintendedtobeusedinternationally.ThisstandardsetsalowerlevelofacceptabilitycomparedtotheANSIstandard. Forcriticalreferents(e.g.safetysymbols),the66%criterionshouldberigorouslyadheredto.Thesestandardsrepresentamajorstartingpointforinternationalharmonizationbut//Kemnitz//(1991)includesareminderoftheconsiderabletimeandeffortrequiredtounderstandanduseanyofthesestandards.      [   [24O  3  .  1  .  8  3  0    STANDARDPHRASES v [݌(#(# Ќ    X]XXXbTheliteraturecontainsmuchinvestigationintotheroleofstandardizedphrasesfor p warnings,particularlyintheuseof signalwordssuchasDANGERorWARNING.ResearchpreviouslycitedbyGodfreyandcolleagues(1993)indicatedthattheuseofasignalphrasesuchas"GovernmentWarning"significantlyinfluencedwarningdetectiontimes.Consequently,therepetitionofwordsorphrasesthroughgovernmentmandatemaybeanimportantwaytoavoidconfusionandincreaserecognition.OtherstudiesonstandardphrasesareincludedintheMSDSportionofthisreport(Section3.3).   w$   w$<24O  3  .  1  .  8  .  1  3  0 `   Signalwords#XbXXX]# / w$<g݌%!` (#` (# Ќ  0  ANSI//Z535.4//1991entitled ProductSafetySignsandLabelsindicateswhenthefollowingsignalwordsshouldbeused: (#(# 0  Dangerindicatesanimminentlyhazardoussituationwhich,ifnotavoided,willresultin )#& deathorseriousinjury.Thissignalwordistobelimitedtothemostextremesituations. (#(# 0  Warningindicatesapotentiallyhazardoussituationwhich,ifnotavoided,couldresultin +%( deathorseriousinjury. (#(# 0  Cautionindicatesapotentiallyhazardoussituationwhich,ifnotavoided,mayresultinminor -`'* ormoderateinjury.Itmayalsobeusedtoalertagainstunsafepractices. (#(#   X]XXXb  TheimportancethatANSIandtheU.S.militaryattachestosignalwordsmaynotbe  sharedbythepopulationatlarge.Giventhelackoftraining,itisquestionablehowthepublicinterpretsdifferentsignalwords.LeonardandMatthews(1993)usedquestionnaireswithcollegestudenttestsubjectstodeterminehowthepopulationingeneraldifferentiateslevelsofwarnings.Nodifferenceswerefoundinratingsofperceptionofrisktodifferentsignalwords.Further,sizeofthesignalwordandcolorofthesignalwordhadnoeffectonperceptionofrisk.Statementsofconsequencesofdisregardingthewarningsandtypeofrisksituationdidaffectratedperceptionofrisk.Also,circumstancesinwhichthesubjectsmightbeplacedaffectedratingsoflikelihoodofdisregardingwarnings.////  Silverand//Wogalter//(1993)testedthearousaleffectsofsignalwordsoncollegestudentsandfoundthatDANGERconnotedgreaterstrengththanWARNINGandCAUTION.TheresultsfailedtoshowadifferencebetweenWARNINGandCAUTION.Amongotherwordstested,DEADLYwasseenashavingthestrongestarousalconnotation,andNOTEtheleast.Fromalonglistof84terms,a"short"listof20signalwordswasdevelopedbasedonunderstandability,lowvariability.shortnessofword,andfrequencyofuse.Theauthorssuggestedthatanexpandedlistofsignalwordsmightalleviatepotentialproblemsofhabituationfromoveruseofthecurrentlyrecommendedterms.#XbXXX]#X]XXXbԀThisresearchwasbuiltuponsimilar,earlierwork(//Wogalter//andSilver,  1990).////  Althoughmanyorganizationshaveguidelinesforthedeterminationofwhatsignalwordsaretobeusedwithspecifichazards,theseareusuallyunknowntothepublic.#XbXXX]# #X]XXXbԀLeonard,Hill,and  Karens(1989)developedinformationabouthowthegeneralpublicperceivesthedegreeofdangerrepresentedbysignalwordsinwarnings.#XbXXX]#X]XXXbԀFor15itemsthathadbeenratedfortheseriousnessof p risk,288subjectswereaskedtoindicatewhichsignalwordtheywouldusetoinformothersofthehazard.SignalwordsthathadbeenfoundtoratehighinseriousnessbyLeonard,Karens,andSchneider(1988)tendedtobeusedmorewithitemsratedashigherrisks.Differenceswerefoundamongagegroupswitholdersubjectsusingsignalwordsthatcarriedmoreseriousconnotations#XbXXX]#X]XXXb. _!0 ////  Therearetwosystemsforcategorizingflammableliquids,onepromulgatedbytheNFPAin1986(NFPA321)andtheotherfoundintheFederalHazardousSubstancesAct,16CFR1500.Implicitinbothsystemsisthatconsumerswillrecognizethedegreeofhazardassociatedwiththeterms flammableand combustible.Main,//Rhoades//,and//Frantz//(1994)showed46driversofstepvanstwophotosoflighterfluidthatwereidenticalexceptonecontainedthewarning flammableandtheother combustible.Thesubjectswereaskedwhichpresentedthegreatestfirehazard.Only23.9%ofthesubjectscorrectlyidentifiedthefluidmarked flammableasthegreaterfirehazardwhile47.8%incorrectlyidentified combustibleaspresentingthegreaterAsecondevaluationinthesamestudylookedathowwellsubjectscouldusethegenericflamesymboltochoosecorrectspecificactionstotakewithanextremelyflammableadhesive.Lessthan40%ofthesubjectsreportedthatvapor//flammability//wasareasontoventilateaworkareaandonly4outof100specificallymentionedthattheywouldavoidusingtheproductnearpilotlights, -`'* themostfrequentlydocumentedcauseofaccidentswiththeseadhesivesandthecauseofabanbytheCPSC.Thisresearchindicatesthattheterminologyandsymbolsoftwocurrentsystemsdonotconveycriticalinformationtoconsumers,accordingtoMain,//Rhoades//,and//Frantz//.#XbXXX]#X]XXXbԀThe  CPSCbanwasinitiatedduetoastudybyNelson(1976)thatfoundtherequiredlabelingwasnoteffective.TheCPSCconcludedthattherewasnotalabelingschemewhichwouldachievethedesiredoutcomeandmovedtoaban.#XbXXX]L#X]XXXb   ////  //Wogalter//etal.(1992)investigatedtheinfluenceofwarningsignalwordsandasignalicononperceptionsofhazardforconsumerproducts.Underthepretextofamarketingresearchstudy,90highschoolandcollegestudentsratedproductlabelsonvariablessuchasproductfamiliarity,frequencyofuse,andperceivedhazard.Resultsshowedthatthepresenceofasignalwordincreasedperceivedhazardcomparedtoitsabsence.Betweenextremeterms(e.g.NOTEandDANGER),significantdifferenceswerenoted,butnotbetweentermsusuallyrecommendedinwarningdesignguidelines.Thepresenceofthesignaliconhadnosignificanteffectonhazardperception.////  ThestandardwisdomofwhetherfourcomponentsareneededforwarningswasevaluatedbyWogalter,Desaulniers,andGodfrey(1993).Fourstatementsignscontainedasignalword,ahazardstatement,aconsequencestatement,andaninstructionstatement,asgenerallyrecommended.Fouradditionalthreestatementsigns,eachwithadifferentcomponentsystematicallyremovedfromthefourstatementsign,wereused,foratotalof5signsforeachhazardsituation.TheresultsofExperiments1and2indicatedthatremovingcontentstatementsreducedperceivedeffectiveness.Hazardandinstructionstatementswerethemostimportant,showingthegreatestdecreaseineffectivenesswhendeleted.Signsforthemosthazardoussituationswereperceivedasthemosteffectivewarnings.Otherresultssuggestedthatthedeletionofredundantstatements,particularlysignalwords,hadlessinfluenceoneffectiveness.Thehazardstatementshowedthelowestredundancyconsistentwithitproducingthegreatesteffectivenessdecrementwhendeleted.////  Ley(1995)lookedateffectivenessoflabelstatementsfortheAustraliangovernment.Thisworkwasprovidedtothe//IOMC//aspartoftheircallonchemicalhazardcommunication.Thestudyindicatedthatthefollowingsignalwordswerenotwellunderstood:combustible,flammable,hazardous,irritant,lethal,andtoxic. #XbXXX]#   w$   w$v"24O  3  .  1  .  8  .  2  3  0    Standardizationoflabelformatw$v""݌&"(#(# Ќ  i"   TheEnvironmentalProtectionAgencyhaswrestledwithwhetherstandardizingpesticidelabelsisaneffectivecourse(EPA,1986).Aftermeetingwithnumerousstakeholdergroupsandestablishinganadvisorycommittee,threeargumentswereconsideredagainststandardization:0  1.0` (#(#Thevarietyofshapesandsizesofcontainersandthevaryingaudiencesreducedthe +%( practicality; ` (#` (#   2.0 ` standardizationwouldeliminateanyopportunityformarketingindividualityby -`'* manufacturers;and ` (#` (#   3.0 ` standardizedlabelswouldallow(orevenpromote)userstosimplyskipoverthe  informationthattheythoughtwasnotimportant. ` (#` (# ////Theargumentsforstandardizationoutweighedtheconcernsabove.Standardformatswereconsideredusefultotheusertoreadilyfindinformationregardingprecautions,practicaltreatment,registrationnumber,etc.Standardizedformatsalsoallowtheusertocompareproducts,uses,risks,etc.andtodeterminewhetherfortheirsituationthebenefitsoutweightherisks.Standardizedlabelswereseenasgreatlyfacilitatingtrainingefforts.TheEPAreviewersagreedthatstandardizationshouldbedonewithincategoriesofpesticideproducts,ratherthanonestandardlabel.Forinstance,thereshouldbeoneformatforagriculturalpesticidesandanotherforhomegardenproducts.Finally,theelementsinastandardizedlabelshouldbeinthelogicalsequencesothatifapersonfollowedeachstepastheyreadthatstep,theywouldbeusingthepesticidesafelyandcorrectly.ThisiswhattheEPAsconsultantcalledthe hierarchyofinformationwiththePrimaryVisualPanelatthetopofthehierarchy(U.S.EPA,1987).////  X]XXXbTheEPA(1986)receivedanoverwhelmingresponsefromstakeholdersandexpertsintheir  evaluationofpesticidelabelingthattherequiredphrase, Keepoutofthereachofchildrenhadnoimpactbecauseofoveruse.Itwasconsideredtheleastreadsectionofthelabels.#XbXXX];+#ԀAfollowup  reportthenextyear(U.S.EPA,1987)recommendeddiscontinuingthechildwarningandreplacingitwithfreshstatementsthatmightberotatedeverytwotothreeyearstoattractattention.   [   [-24O  3  .  1  .  9  3  0    ADDITIONALRESEARCHNEEDS -[--݌(#(# Ќ  //Dorris//andPurswell(1978)presentedalistofneededresearch,including:  1.0 ` Optimumamountofinformationtobepresented;R` (#` (#   2.0 ` Symbolicversusverbalwarningeffectiveness;q B` (#` (#   3.0 ` Needforanappropriatemethodologyforstudyingbehavior;anda!2` (#` (#   4.0 ` Needtounderstandthefactorsthatinfluenceresponsestowarning.Q""` (#` (#   //Purswell//,//Krenek//,andDorris(1993)builtonthepreviouslistandincludedtheobservationthattheconceptualmodelofthewarningprocessdevelopedbyMillerandLehto(1986)alsoneedstoberefinedtohighlightalloftheadditionalvariablesthatneedtobestudied.Purswellandcolleagueshighlightedthefollowingareasformoreresearch:  Stimulus:EnergyLevelandContact  '!$ 0  Mostusersofaproductwillnotreadthewarninginformationbecauseofawelldocumented filteringprocess. (#(# 0  Thefollowingvariablesaffectfiltering: (#(#   1.0 ` Informationoverload,r&)` (#` (#    ` a.0 Warninglistsfrequentlyincludetoomanyitems;-b'* (# (#    ` b.0 Contentsofasinglewarningmaybetooextensive;/ (# (#    ` c.0 Theremaybetoomanyindividualwarningsplacedinthefieldofview;and (# (#    ` d.0 Theremaybeothernonwarningstimuli. (# (#   2.0 ` FaultyriskassessmentTheamountofriskinformationprovideddoesnot  significantlyinfluencethesubjectiveratingofhazardperceived. ` (#` (#   3.0 ` Benignexperienceversusawarning//Karnes//etal.(1986)hypothesizethatbeing   regularlyexposedtoawarningaboutahazardwhileatthesametime,thehazarddoesnotcauseaninjury,filteringwilltakeplace. ` (#` (#   ComprehensionofWarnings  1.0 ` Themeaningfulnessof signalwords` ` (#` (#   2.0 ` Thereadingcomprehensionlevelrequiredtounderstandthewarningparticularly, P  thetradeoffbetweentheuseofasmallernumberofwordswithamoreexactmeaningversusalargernumberofsmallerwordswithlessexactmeanings. ` (#` (#   3.0 ` ThemeaningfulnessofsymbolsO ` (#` (#   4.0 ` Themeaningfulnessofwarningasafunctionofthetaskbeingperformedthisisthe ? effectivenessofwarningswhenpresentedinthecontextofinstructionsascomparedtopresentingthewarningsinaseparatelistwithoutthetruecontextofperformingsomeoperation. ` (#` (# 0  WarningandMemory (#(# 0  Fewwarningsarestoredinlongtermmemoryiftheyaremorethansixorsevenlineslongoraddressmorethanthisnumberofhazardsinusingaproduct.Moreresearchisneededtobetterunderstandthereasonsforthelackoflongtermrecall. (#(# [ 0  TheDecisionMakingProcess (#(# 0  Purswell,Krenek,andDorrisconcludetheirlistofneededresearchwiththisitem,whichtheyidentifyasthemostimportant.Theyhypothesizeathresholdofperceivedprobabilityofaninjurywhichmustbereachedbeforeapersonsbehaviorwillbeinfluencedbyriskinformation.Thisthresholdmayberelativelyhigh,i.e.1/100beforemostpersonswillrespondtoknowledgeabouthazardsaspresentedinawarning.Sincefewproductscarrysuchrisks,mostwarningsgounheeded.//Purswell//etal.citeGodfreyetal.(1985)conceptof costofcompliance.Theremaybebenefitsofnoncompliancethatareofsignificantimportancetotheindividual.Consequently,individualsappeartoconstructsometypeofutilityfunctionformakingthetradeoff.Thisisanotherimportantareaforresearch. (#(#    s   sf@24O  3  .  2  3  0    TRAININGsf@@݌& #(#(# Ќ  X@   [   [jA24O  3  .  2  .  1  3  0    SUMMARYOFFINDINGSANDRECOMMENDATIONS ]A[jAA݌("%(#(# Ќ  0  Thefollowinginformationrepresentsasummaryofthegeneralareasofconsensusinthe *$' literatureontraining.Thesummaryalsoincludeswhattheauthorsofthisreport. +%((#(# Y Z?  Z?CY1  .0    Reducedworkerinjuryrateswerefoundaftertheintroductionof//worksite//firstaidtraining -`'* programsinseveralstudies,suggestingthatthiskindofinstructionincreasesconsciousnesstojobsafetyconcerns.Z?CCYی (#(# Ќ  Z?  Z?UEY2  .0    Thereappearstobeaclearlinkbetweentrainingandtheestablishmentofhealthyandsafe  workingconditions.Thestudiesreviewedinonemetaanalysiswere nearunanimousinthebenefitsthattrainingcanachievesuchasincreasedhazardawarenessandsafeworkplacepractices.Z?UEpEYی (#(# Ќ  Z?  Z?=GY3  .0    EvaluationoftheU.S.HazardCommunicationStandardinselectedmanufacturingplants   showedthatthemostimportantvariableindeterminingbettercompliancewasexplicitsupportfortheprogrambyplantlevelmanagement.Z?=GXGYی (#(# Ќ  Z?  Z?HY4  .0    Inprogramspreparinghealthandsafetyprofessionals(atbothundergraduateandgraduate p  levels),verylittlecourseworkpreparesthestudentsfortheirrolesascommunicatorsortrainers.X]XXXbZ?HHYیP (#(# Ќ  Z?  Z?JY5  .0    #XbXXX]J#Literacyandlanguageissueshavenotbeensubstantiallyexploredregardinghazard o@  communicationandthereforeshouldbeevaluated.Z?JJYی (#(# Ќ    [   [L24O  3  .  2  .  2  3  0    GENERALCONSIDERATIONS L [L:L݌A(#(# Ќ    Hazardcommunicationisaccomplishedthroughaprocessthatincludesseveraltoolsfortransferringinformationandinfluencingbehavior.Laugherand//Brelsford//(1993)notethat numerousstudieshaveshownthegreatertheaprioriperceptionofhazardousness,themorelikelypeoplewilllookforandreadawarning,andthemorelikelytheywillcomplybytakingsafetyprecautions(Donnerand//Brelsford//,1988;Godfreyetal,1983;//Friedmann//,1988;//LaRue//andCohen,1987;Leonardetal.,1986;and//Wogalter//etal.).Trainingprogramsareoneofthecomponentsofahazardcommunicationprogramthatcanprimetheworkerspumpsothattheyarereceptivetotheimportantmessagesfromothersourcesofcommunication.      [   [P24O  3  .  2  .  3  3  0    //NIOSH//LITERATUREREVIEW P [PP݌b(#(# Ќ    In1995,a//NIOSH//publishedareportentitled AssessingOccupationalSafetyandHealthTraining:ALiteratureReview(Cohenand//Colligan//,1995)Thefollowingistheauthorsabstractinitsentirety.0  Morethan100OccupationalSafetyandHealthAdministration(OSHA)standardsforhazardcontrolintheworkplacecontainrequirementsfortrainingaimedatreducingriskfactorsforinjuryordisease,andotherslimitcertainjobstopersonsdeemedcompetentbyvirtueofspecialtraining.Aliteraturereviewwasundertakentoassessthemeritsofsuchtrainingrulesinachievingthisobjectiveandtosortoutfactorsofconsequence.Thereviewfocusedheavilyonreportsofstudieswheretrainingwasusedasaninterventionefforttoreduceriskofworkrelatedinjuryanddisease.Thisliteratureyieldedmuchevidencetoshowthemeritsoftraining.However,thefindingshadtobetemperedbecausea)theworkdidnotaddressOSHArulesperseb)knowledgegainandsafebehaviormeasuresasopposedtoactualinjury/diseaseindicatorswereusedinmanyevaluations,andc)thetrainingwasinsomeinstancescoupledwithotherformsofinterventionsoastomakeattributiondifficult.Reportsfromselectsurveysandinvestigationsofworkerinjuriesandworkplace -b'* facilitieswerealsoaccessedandgavemixedresultswithregardstotherulethattrainingdeficitsmayhaveplayedasacontributingfactor.Regardingotherworkplacetrainingactivitiesofrelevance,reducedworkerinjuryrateswerefoundaftertheintroductionof//worksite//firstaidtrainingprogramsinseveralstudies,suggestingthatthiskindofinstructionincreasesconsciousnesstojobsafetyconcerns.Resultsfromanalyzingthehazardcontrolprogramsofcompanieswithexemplarysafetyandhealthperformancerecordsfoundmoretimedevotedtotrainingandthegreaterinvolvementofsupervisorsinsucheffortstobeimportcharacteristics. (#(# 0  Otherfactorsfoundtoinfluencethetrainingprocessanditsimpactsattheactual//jobsite//emergedfromthereview,thoughthedatainsomecaseslimitedthestatementsthatcouldbemadeastotheirsignificance.Referenceismadetosizeoftraininggroup,lengthandfrequencyoftraining,mannerofinstructor,trainingcredentials,goalsetting,feedback,managementcommitmentamongothers.TheOSHAvoluntarytrainingguidelinesaredescribedalongwithillustrationsfromthereportstoshowhowthevarioustrainingguidelinesaredescribedalongwithillustrationsfromthereportstoshowhowthevariousstepscontainedwithinthemcanbemetinrealisticwaysandhavemeritinframinganeffectiveprogram.Basedontheliteraturereview,followoneffortstoaddressoutstandingissuesandneedsregardingeffectiveoccupationalsafetyandhealthtrainingarenoted. (#(#   Cohenand//Colligan//lookedattraininginterventioneffortsdesignedto enhanceworkersknowledgeofworkplacehazards,affectbehaviorchangestoensurecompliancewithsafeworkpracticesorpromptotheractionsaimedatreducingtheriskofoccupationalinjuryordisease.Thereportestablishedaclearlinkbetweentrainingandtheestablishmentofhealthyandsafeworkingconditions.Theauthorsnotedthatthestudiesreviewedwere nearunanimousinthebenefitsthattrainingcanachievesuchasincreasedhazardawarenessandsafeworkplacepractices.Twocaveatsmustbenoted.First,thetypesoftrainingprogramsthatwereevaluatedweretargetedtothespecificworkplace,notnecessarilytoanOSHAtrainingrequirementandsecond,theevaluationmethodologiescouldnotpinpointtheprecisebenefitthatthe trainingperseaccomplishedintheoverallhealthandsafetyprograms.////  Theauthorsnotedseveralothervariablesthatappearcriticaltoaccomplishingtheoverarchinggoalofhealthyandsafeworkplacesincluding, Managementsroleandsupportofsafetytraininganditstransfertothe//jobsite//,settinggoals,andprovidingfeedbacktomotivateuseoftheknowledgegained,andofferingincentives,rewardsforreinforcingsafeperformance(Cohenand//Colligan//,1995).Theseareallcriticalpiecesinensuringasuccessfulprogram.InanevaluationoftheimplementationoftheHazardCommunicationStandardinmanufacturingplantsinMaryland,Sattler(1990)foundthatthemostimportantvariableindeterminingbettercompliancewasexplicitsupportfortheprogrambyplantlevelmanagement.////  SomeofthestudiesreviewedbyCohenand//Colligan//didnotevaluatethequalityofthetraining,merelythepresenceoftraining.Thedatawerenotasstrongindefiningthefactorsthatcontributetoaneffectivetraining. -`'*    [   [e24O  3  .  2  .  4  3  0    RIGHTTOUNDERSTAND[ef݌(#(# Ќ  e   TheLaborOccupationalHealthProgram(//LOHP//)attheUniversityofCaliforniaatBerkeley,producedX%XXXbTheRighttoUnderstand:LinkingLiteracytoHealthandSafetyTraining,#XbXXX%g#Ԁaspartofa   projectfundedbytheNIEHS(//Szudy//andArroyo,1994).Thedocumentisapeerreviewed,wellreferenced,andpracticalworkthatprovidesguidanceforhealthandsafetytrainers.Itemphasizesthechallengesintrainingadultlearnerswhomaybeilliterateorhavereadingdifficulties(becauseofliteracyissuesorEnglishasasecondlanguage).Ittakesintoaccountbasicprincipalsofadulteducation,healtheducation,andtrainingmethodologiesthathavebeenexpresslyadaptedforhealthandsafetysubjectmatter.Additionally,itprovidesareviewofliteracytermsandstatistics,includingthefactthatanestimatedonethirdoftheU.S.workforcereadsatorbelowtheeighthgradereadinglevel.////  Thebookhasachapterspecificallydevotedtocreatingmaterialsthatareeasytoread,visuallyappealing,andillustratedtohelpexplainthetext.Thereisachapter,aswell,onfieldtestingmaterials.Thebookcontainsseveralgeneralrecommendationsthatareappropriateforconsiderationinimprovinghazardcommunicationtraining:  1.0 ` Maketechnicalhealthandsafetyinformationmorereadablefor everyone ,including  thosewithreadingandwritingproblems; p ` (#` (#   2.0 ` Adapttrainingmethodstorelylessonworkersreadingskills;and` (#` (#   3.0 ` Developnewmethodsofassessingwhatworkerslearnfromtraining.` (#` (#   TheperformancebasisfortheHazardCommunicationStandardmandatesthattheemployerisresponsibleforinformingworkersaboutchemicalhazardsonthejob.Thisincludestrainingworkerswhocannotreadorwhohavepoorreadingskills.Afterevaluatingthereadabilitylevelsof25healthandsafetymaterialspreparedbygovernmentagencies,unions,workereducators,andprivatesafetycompanies,//LOHP//discoveredthattheaveragereadabilitylevelwasacollegereadinglevel.Manyhealthandsafetymaterialsarewrittenbytechnicalstaffwhohaveneverreceivedanytrainingonpreparingdocumentsorcommunicationtoolsfor any audience. O"    Asideissueisthefactthatinprogramspreparinghealthandsafetyprofessionals(atbothundergraduateandgraduatelevels),verylittle//coursework//preparesthestudentsfortheirrolesascommunicatorsortrainers.ThishasbeencorroboratedinthePublicHealthServiceReportonthePreparationoftheEnvironmentalHealthWorkforce(1991).////  //Wallerstein//haswrittenextensivelyonworkertrainingandinaspecialissueoftheX%XXXbAmerican ("% JournalofIndustrialMedicine:EmpowermentApproachestoWorkerHealthandSafetyEducation(1992),#XbXXX%=s#sheoutlinesissuesassociatedwithlowliteracyandEnglishasasecondlanguageand *$' encouragestheuseofparticipatorytrainingtechniquestoovercomethesechallenges.Literacyandlanguageissueshavenotbeensubstantiallyexploredregardinghazardcommunicationandthereforeshouldbeevaluated. -`'*    [   [ku24O  3  .  2  .  5  3  0    ADDITIONALRESEARCHNEEDS ^u [kuu݌(#(# Ќ  SuggestionsweremadebyCohenand//Colligan//forfurtherworkintheareaofhealthandsafetytraining:1.0  UndertakingstudiestoascertainhowindustryisrespondingtoOSHAtrainingrulesandthe   qualityofsuchefforts.Themajordatasetusedinthe//NIOSH//reviewwereresearcherdirectedefforts,andforthatreasonwerenotthenorm.Focusingtheeffortsonthemostprevalenttypesofinjuriesandillnessesandselectingindustriesandworkoperationswheretheyaremostrecurrentwouldbeideal.Differencesinhowthemandatedtrainingrulesweremetatthevarioussitesselectedandlinkagesbetweenthetrainingundertakenandspecificinjuryanddiseaseriskfactorswouldbeanalyzed.TheextenttowhichtheoperantpracticesfollowedOSHAtrainingguidelines,resultantexperiencescouldofferanimportantreferenceingaugingtheirutility. (#(# 2.0  Conductingindepthstudiesoftrainingpracticesandtheirinterrelationswithotherelements O  inanestablishmentshazardcontrolprogram.Directingthiseffortatcompaniesshowingexemplarysafetyandhealthrecordscouldofferprogrammodelsforeffectivetrainingwhichcanbestcomplementorenhanceotherworkplacemeasuresaimedatmaximizingriskmanagement. (#(# 3.0  Usingcasecontrolorcohortstudiestocomparedifferencesintheleveloftrainingof  workersinjuredorafflictedbyoccupationaldiseaseagainstthosenotsoaffected.Theintentherewouldbetogetabetterassessmentofhowtrainingdeficitscanleadtosuchproblems.Suchananalysiswouldrequiremeasurestoseparateoutmanynontrainingfactorsthatcouldalsoberesponsibleforapparentdifferentialsinthesecases. (#(# 4.0  Conveneoneormoreworkshopstodiscussissuesconcernedwiththeeffectivenessof  //worksite//occupationalhealthandsafetytrainingbothnowandinthefuture.//Invitees//wouldincludeexpertsandpractitionersconversantwithoccupationalsafety/healthtraining,jobskillstraining,healtheducation,organizationalbehavioral,andevaluationsubjectareas.Theworkshopswouldseektopoolideasbearingonthequestionsposedinthe//NIOSH//reportandaddedconcernssuchastheadequacyofcurrentregulatorylanguageinOSHAtrainingrules,futuretrainingchallengesduetochangingworkplacetechnologies,workerdemographics,themeritofmergingdifferentworkplacetrainingdomains(i.e.,occupationalsafetytraining,jobskillstraining,//worksite//healthpromotion),andotherissues. (#(#   s   s24O  3  .  3  3  0    MATERIALSAFETYDATASHEETSs݌&"(#(# Ќ    [   [24O  3  .  3  .  1  3  0    SUMMARYOFFINDINGSANDRECOMMENDATIONS [ǂ݌'!$(#(# Ќ  0  ThefollowingrecommendationsrepresenttheconsensustakenfromtheliteratureonMSDS )#& research,aswellaskeypointstheauthorsofthisreportfeltneededemphasizing. (#(# YZ?  Z?˄Y1  .0    Therehasbeenlittleresearchcomparingformatsofsafetydatasheetsbutwherecomparisons ,p&) havebeenmade,the//ICSC//hasbeenrankedmoreeffectivethantheotherformatstested.Z?˄Yی-`'*(#(# Ќ  Z?  Z?cY2  .0    Onaverage,literateworkersonlyunderstoodabout60%ofthehealthandsafetyinformation / ontheMSDSsassociatedwiththehazardouschemical,inallthreecomprehensibilitystudies.Z?c~Yی (#(# Ќ  Z?  Z?Y3  .0    RecommendationsforcontinuedworkandresearchonMSDSs:Z?Yی(#(# Ќ        و"0 `   FindwaystoimprovethereadabilityandcomprehensibilityofMSDS. و݌` (#` (# Ќ        "0 `   ResearchthewaysinwhichtheformatoftheMSDSinfluencescomprehensibility,   includingtheuseofstandardizedglossaryofterms. ۉ݌ ` (#` (# Ќ        ܊"0 `   Determinetherolesthatlabelingandtrainingcanplayincomprehendingthehealth   andsafetyissuesassociatedwithpotentiallyhazardouschemicalexposuresintheworkplace. ܊݌ ` (#` (# Ќ  Z?  Z?0Y4  .0    Anothersetofrecommendationsincluded:offeringliteracyprogramsinworkplaces; p  incorporatingliteracyprogramsintohealthandsafetytraining;employingavarietyofmethodsforcommunication,inadditiontowrittenmaterials; //buddying//newemployeeswithveteranemployees;andmakinghealthandsafetydocumentsmorereadable.Z?0KYی (#(# Ќ  Z?  Z?bY5  .0    MSDSs,bythemselves,areapoormeansofinformingworkersofhazardstowhichthey _0  maybeexposedforthefollowingreasons:1)muchofthetechnicalinformationhaslittlemeaningtotheaverageworkerandmayevenfrustratetheworkersabilitytoreadotherportionsoftheMSDSsthathaveinformationpertinenttohazardrecognitionandsafepractices;2)informationdepictinghazardousconditions,signsorsymptomsofexposure,andsafehandlingproceduresarewrittengenericallyandworkersmayhavedifficultyseeingtheconnectionbetweentheirownuseofthechemicalandtheinformationontheMSDS;and3)theinformationmaycontaintermstoodifficulttounderstandortoobriefandvaguetoactuallygeneratetheconcernthatworkershouldhaveregardingsafeuseofthechemicals.Z?b}Yی (#(# Ќ  Z?  Z?LY6  .0    Oneexpertpanelreviewestablishedthatonly11%oftheMSDSswerefoundtobeaccurate  inallofthefollowingfourareas:healtheffects,firstaid,personalprotectiveequipment,andexposurelimits.Further,thehealtheffectsdataontheMSDSsfrequentlyareincompleteandthechronicdataareoftenincorrectorlesscompletethantheacutedata.Z?LgYی (#(# Ќ  Z?  Z?vY7  .0    Ofparticularimportancetotheinternationalharmonizationeffort,theoverwhelming ` majorityofthenationsthatrespondedtotheCoordinatingGroupfortheHarmonizationofChemicalClassificationSystems(InterOrganization//Programme//fortheSoundManagementofChemicals,December,1996)reportedtheyrequiredorrecommendedtheuseofSafetyDataSheets.Nearlyallofthosecountrieswithsafetydatasheetsrequirementsstipulatedwhatsectionshadtobeincludedinthesheets.Z?vYی (#(# Ќ   [   [=24O  3  .  3  .  2  3  0    INTRODUCTION [=h݌%!(#(# Ќ   w$   w$D24O  3  .  3  .  2  .  1  3  0    InternationalUseofSafetyDataSheets w$Do݌& #(#(# Ќ    The//IOMC//draftreportresultedfromcontactwith270nationalandinternationalinstitutions.A fewselectedorganizations(//IOMC//,1996)weresentaquestionnairetoascertainmoreinformation.Ofthetennationsthatresponded,allbuttwohadaregulatoryrequirementorrecommendationthatsafetydatasheetsbeused.Sevenoftheeightrequiredorrecommendedspecificsubjectstobecoveredinthedocument.AsofMay,1997Denmark,Finland,France,Germany,Norway,Sweden,Switzerland,theUnitedStates,andtheUnitedKingdomrequiredthe -`'* productionanduseofMSDSbutonlyFinlandrequiredproducerstouseaspecificformat(Phillips,1997).TheCanadianformatiswellknownandiscalledtheWorkplaceHazardousMaterialsInformationSystem(//WHMIS//).Thesystemrepresentsagoodexampleofthedifficultiesfacingattemptsatinternationalharmonization.The//WHMIS//couldbeusedintheUnitedSatesbutthereciprocalisnottrue(West,1991).ThecreationoftheEuropeanEconomicCommunitygavestrongimpetustorequiringpreparationofMSDSforallchemicalsandmixtures(Campbell,1992).//Worksafe//AustraliaadoptedanationalMSDSformatin1989.////  SomeEuropeancountrieshavealreadyadoptedtheANSI//Z400//format.ThissixteensectionCMA/ANSIformatwascoordinatedwithmajorchemicalassociationsofCanada,Europe,andJapaninreachinganagreementonsectioncontentandtitles.In1990,theGeneralConferenceoftheInternationalLaborOrganizationadoptedConvention170andRecommendation177concerningsafetyintheuseofchemicalsatworkandaclassificationschemeusingthe16sectionsoftheANSIformat.TheCommissionofEuropeanCommunitiesinaJune5,1991Directivedelineateditssafetydatasheetformat,whichincludedthesame16sectionheadingsastheANSI//Z400//documentrecommends.Thisdirectivemadeuseofthedocumentmandatoryanditwastobeavailableinappropriateuserlanguages.AsummaryofthedataelementsoftheOSHAHazardCommunication//Sandard//,Canadas//WHMIS//,andthe//EC//DirectivecanbefoundinAnnexAoftheANSI//Z400.1//1993standard.////  AnotherimportantinternationaldevelopmentinhazardcommunicationhasbeenthecreationofInternationalChemicalSafetyCards(//ICSCs//).ThiseffortthroughtheInternational//Programme//onChemicalSafety(//IPCS//)isajointactivityofthreecooperatinginternationalorganizations:theUnitedNationsEnvironmentProgram,theInternational//Labour//Office,andtheWorldHealthOrganization.Theirworkhasbeenfocusedonprovidinginformationtodevelopingcountries.Thegoalsweretodevelopastandardizedformatthatwasshorter(onedoublesidedpage)andlesscomplexthanANSIapproach.Anothergoalwastoidentifystandardphrasesfromtheliteratureandthenincorporatethemintothecard(//Niemeier//,1997).////  TheInternationalChemicalSafetyCardssummarizesessentialhealthandsafetyinformationonchemicalsfortheiruseonthe shopfloorlevelbyworkersandemployersinfactories,agriculture,andconstruction.Theyareverifiedandpeerreviewedbyinternationallyrecognizedexpertsandcouldserveastheprincipleinformationsourceinlessdevelopedareasorinsmallandmediumsizeenterprises.The//IPCS//intendstocreatenearly2000cardsinthenextsixyears(//NIOSH//,1997).Ofparticularimportancefortheharmonizationefforts,the//IPCS//hascomputerizedapproximately1200standardphrasesandwillbetestingaWindows95compatiblesoftwarepackagesoonthatallowsdesignerstocreatethesecardsusingstandardphrasesandthentranslatethemintoanyoftheavailablelanguagesautomatically.The//ICSCs//andthesoftwarepackageshouldbeavailableontheInternetsoon(//Niemeier//,1997).////  Duetoconcernsaboutliteracylevelsamongitsworkforce,DowChemical,in1989,beganusingiconsinsteadofstandardphrasestoinformworkersofhazards.//NIOSH//haspickedupthissystemsandhasconvertedthestandardphrasesinitsdatabasetoicons.Thesewillbetestedon -`'* workersemployedaspaintersthroughtheCentertoProtectWorkersRights.TheywilluseaCDRommultimediaapproachprovidingthestandardphrasesonascreenalongwiththeassociatediconwhilepronouncingthephrasesaspartoftraining(//Niemeier//,1997).  TheMSDSisoneoftheprimetoolsforinformationtransferintheimplementationofhazardcommunicationprograms.AstudycontractedbyU.S.OSHA,evaluatedthecomprehensibilityofMSDStoworkers,theprimeaudienceforhazardcommunication.Thestudy(//Kolp//etal.1993)wasdonepriortothedevelopmentofthenewANSI//Z400.1//1993StandardforHazardousIndustrial   ChemicalsMaterialSafetyDataSheetPreparation(hereafterANSI//Z400.1//)andexaminedthe   comprehensibilityofasampleofMSDSstoagroupofabout100unionizedworkersinmanufacturingindustrieslocatedinthestateofMaryland.////  Thepurposeofthestudywastoassesstheabilityoftheworkerstounderstand,foreachMSDSreviewed,therouteofentryofthesubstance,thetypeofhealthhazardpresented,whatneedstobedonetoavoidorprotectagainstthehazard,andwheretogoifadditionalhelpisnecessary.AliteraturereviewindicatedthatthereadabilityofanMSDS,i.e.thereadinggradelevelatwhichtheMSDSiswritten,isanimportantfactorindeterminingthecomprehensibilityofanMSDS.Tocontrolforthisvariable,theinvestigatorschoseMSDSswiththeaveragereadabilitylevelataboutatwelfthgradelevel,asdeterminedbyevaluatingthereadabilityof100randomlyselectedMSDSs.////  Basedontherecommendationsofareadingexpertinhealthmessages(//Freimuth//1979),theSMOGGradingFormulawasappliedinthestudytodeterminethereadabilitylevelsoftheMSDSs.Thisformulationisasimpleandfasttestbasedontheproportionofpolysyllabicwordsinadocument.TheSMOGformula,developedbyMcLaughlin,wastheassessmenttoolofchoiceselectedbytheOfficeofCancerCommunication,attheNationalCancerInstitute.TheMSDSformatcouldnotbecontrolledbecausetheMSDSwererandomlyselected.Amongonegroupofworkers,anInternationalChemicalSafetyCard(//ICSC//)wasalsotested.Whentestedforreadability,the//ICSC//readatan11thgradereadinglevel,andontheaveragewasbetterunderstoodbyworkers.////  Beforereportingtheresults,itmustbenotedthattheworkerswhovolunteeredforthisstudyunderstoodthatitreliedonreadingcomprehension.Thispresentedaselectionbiasandcreatedabestcasescenariofortheresults.WorkerswithreadingdifficultiesorEnglishastheirsecondlanguagewouldnothavevolunteeredforthestudy.(Approximately22%oftheUSadultpopulationisfunctionallyilliterate.)Nonetheless,itwasfoundthatonaveragetheworkersonlyunderstoodabout60%ofthehealthandsafetyinformationontheMSDSsassociatedwiththehazardouschemical.Thoughthescopeofthestudywaslimited,theresultswereacausetoreconsidertheroleofMSDSasacommunicationtool.InreportingthisstudytoOSHA,theinvestigatorsmadethefollowingrecommendationsforcontinuedworkandresearchonMSDSs:      w"0 `   FindwaystoimprovethereadabilityofMSDS. w݌)#&` (#` (# Ќ        <"0 `   ResearchthewaysinwhichtheformatoftheMSDSinfluencescomprehensibility, <c݌*$'` (#` (# Ќ     ` includingtheuseofstandardizedglossaryofterms.      p"0 `   Determinetherolesthatlabelingandhealthandsafetytrainingcanplayin @ 0@ showing @  ,p&) comprehendingMSDSs,withparticularemphasisonhowbesttointegratetheuse -`'* ofMSDSsintohealthandsafetytrainingtocommunicatethehazardsassociatedwithchemicalsubstances. p݌ ` (#` (# Ќ  ////  Anotherstudywasreportedin1990,bythePrintingIndustriesofAmerica(PIA,1990).ThestudyinvolvedthecomprehensibilityofMSDStoMasterPrinters,whohadanaverageof13.9yearsofformaleducation,orapproximatelytwoyearsbeyondhighschool.Inthisstudy,27MSDSwereselectedandanalyzedforreadinglevelsusingasoftwareprogram,findingtheaveragereadinggradelevelof14.Theinvestigatorsfoundthatemployeeswith15yearsofeducationormoreunderstoodonly66.2%.Basedontheseresults,itwassuggestedthatyearsofformaleducationdoesnotsignificantlyincreasecomprehensionofMSDS.////  Atthetime,thePIAmadethesuggestionthatastandardizedformatbeusedforMSDS,aswellasstandardizedcolorcodedsectionsforeasieraccesstoemergencyinformation,andstandardizedsignsandsymbolsforvitalinformation.TheyfurthersuggestedthatMSDSbeproducedinotherlanguages(thisisnotrequiredbyOSHA)andbewrittenatareadinglevelnohigherthantwelfthgrade.////  InaCanadianjournalonoccupationalsafetyandhealth,//Nore//(1990)expressedtheneedtobealerttotherelationshipbetweenliteracyandhazardcommunication.//Nore//citedaCanadianstudyfindingthatsevenpercentofageneraladultpopulationcouldnotuseprinteddocumentscommonlyfoundatwork,home,orinthecommunity,andthat20%couldnotreadtheinstructionsthatcomewithaspirin.Lackofreadingskillmakesfull,safeandproductiveworkplaceparticipationdifficultwhenworkersaredependentonthewrittenwordtocommunicate.//Nore//hadseveralrecommendations:offeringliteracyprogramsinworkplaces;incorporatingliteracyprogramsintohealthandsafetytraining;usingavarietyofmethodsforcommunicating,inadditiontowrittenmaterials; //buddying//newemployeeswithveteranemployees;andmakinghealthandsafetydocumentsmorereadable.////  Inanarticleaboutfunctionallyilliterateworkers,//Samways//(1988)raisedtheconcernthataccesstoMSDSsmaysatisfytheemployeesrighttoknow,howeveritmaynotsatisfytheirrighttounderstand.//Samways//suggestedtheuseofsimplified//factsheets//.//Samways//alsosuggestedtrainingmethodologiesthatincorporatevarietyinthepresentationmedia,thatprovideanopportunitytoaskquestions,andprovidetraininginlanguagesresponsivetotheaudienceslanguageneeds.////  AmajorefforttoimprovethequalityoftheMSDSisrepresentedbytheconsensusbuiltdevelopmentoftheANSI//Z400.1//StandardforthedevelopmentofMSDSstobeusedunder industrialoccupationalconditions.However,itrecognizesinitssectiontitled AudiencesforMSDSsthatthefollowingindividualsmayuseMSDSsastheirsourceofinformation:communitymember,emergencyresponder,employee,employer,environmentalprofessional,medicalprofessions,andoccupationalhealthandsafetyprofessional. -`'* ThefollowingisextracteddirectlyfromSection3.2oftheANSI//Z400.1////Standard:// ؖXXb553.2  ReadingLevelandComprehension   #XbXؖ#Oneofthegreatestchallengesinpreparingamaterialsafetydatasheetiswritingsothatvariousaudiencescanreadandunderstandtheinformation.Readinglevelsofusersvarywidely.Thetargetaudiencesrangefromanuntrainedpersonneedinggeneralinformationtoahighlytrainedprofessional.TheinformationbeingconveyedisoftenverytechnicalandmustbecompleteenoughforthespecialistandyetbeunderstandableforthelesstrainedMSDSuser.Anadditionalchallengeisthatthetargetaudienceschangefromsectiontosection.Therefore,thereadinglevelshouldchangeaswelltoadapttothetargetaudiences.Wordchoiceandsentencestructuregreatlyeffectreadinglevelandcomprehension.Thefollowing rulesofthumbmaybehelpfultoimprovethereadabilityandcomprehensibilityoftheMSDS.0  Keepsentenceshortanddirect.Usenomorethantwosubordinateclauses.Usetheactivevoiceasmuchaspossible. (#(# q3:8;+'h|  `EHhq m(#m(#(#(#  Instructions (#(#m(#m(#aremorelikelytobefollowedifconsequencesaredescribed.0  Sentencesthatincludealongstringofeffectsorotheritemscanbemadeclearerbyputtingthemintolist. (#(# q3;8;+'h| _  `EF g  P~ q (#(#(#(# (#(#(#(#0  Useshortwordsofoneortwosyllablesasmuchaspossible.Choosecommonlyused,familiarwords,butavoidcolloquialismsandslang. (#(# 0  Useonlycommonabbreviationsandacronyms,andthengivetheirdefinitionassoonaftertheirfirstuseaspossible.Occasionally,however,anabbreviationoracronymmaybesofamiliartointendedaudiencesthatitmaybeusedwithoutadefinition.Infact,somemaybemorefamiliarthanthefullname(e.g.,OSHA,EPA,SARA,$F,$C,//TLV//,andTWA)-v'*(#(# Ї0  Avoidtechnicallanguageandjargonexceptforinformationessentialtoappropriatelytrainedindividuals.Whentechnicallanguageisnecessaryinsectionstargetedfornontechnicalaudiences,itisadvisabletoalsoincludealesstechnicalexplanation.Refertotheglossaryforexamplesofsomeusefulalternativesfortechnicalwords. (#(# 0  ResponsestoMSDSsfromworkersandcustomerscanbeveryusefulindeterminingcomprehensionandthereadabilityofMSDSs. (#(# 0  Theuseofaconsistentstatementorphraseisbeginningtogainpopularity,especiallywheretranslationbecomesaconsideration. (#(# @@` @  TheANSI//Z400.1//StandarddoesnotprovideastandardizedformatbutratherstandardizesthestepstobetakeninpreparinganMSDSandcreatesastructure(16sectionheadings)inwhichtowork.Thestandardiscurrentlyunderitsfirstrevision.NostudyhaseverbeendonetodeterminewhetherMSDSspreparedfollowingtheANSI//Z400.1//recommendationsweremoreorlesscomprehensiblethantypicalMSDSs.  ThisraisesanimportantquestionregardingtheeffectivenessofthefollowingrecommendationsmadetoOSHAbytheNationalAdvisoryCommitteeonOccupationalSafetyandHealthHazComWorkgroup:0  OSHAshouldendorsetheorderandsectiontitlesasdescribedinANSI//Z400.1//1993,and  stronglysupporttheuseofotherrecommendationsfromthisvoluntarystandard.ThiscanbedonethroughreleaseofanOSHAmemorandum,aninterpretiveletter,interpretivenoticeand/orlistingoftheorderandsectiontitlesasanonmandatoryappendixtothe//HCS//(HazardCommunicationStandard).`(#(#   Asmentionedbefore,theANSI//Z400.1//StandardprovidesguidanceforthepreparerofMSDSstoincludetheinformationinastandardizedsequenceandgivesguidanceonaddressingreadabilityissues.However,accordingtoamemberoftheSocietyforChemicalHazardCommunicationwhoisnowprovidingnationaltrainingonMSDSspreparationusingANSI//Z400.1//,ifseveralprofessionalswereaskedtousethestandardtodevelopanMSDSonaspecificchemical,theresultingMSDSscouldlookquitedifferent(//Ignatowski//,1997).////  AselectionofstandardphrasesrecommendedintheANSI//Z400.1//wereformallyevaluatedinastudycommissionedbytheChemicalManufacturersAssociation(//Lehto//,1993).Inthisstudy,63standardphraseswereevaluatedonthebasisofsyntacticanalysis(usingcomputerizedreadabilitymethods);semanticanalysis(determiningtheuseandunderstandabilityofindividualwords);andaStructuredFocusGroupanalysisdeterminingthecomprehensibilitybyagroupofchemicalworkers.Itwasfoundthatthechemicalworkersratedtheirunderstandingofnearlyallofthephrasessomewherebetween understandcompletelyand understandmostly.// -`'* ////////  Inthesecondstudy,workersenrolledinastudyinwhichtheyknewtheywouldbetestedonreadingcomprehension.Theireducationlevelswereasfollows:  EDUCATIONLEVEL h 0    #OFWORKERS (#(#    ` eighthgrade   h      p 0      ` twelfth   h      p 25      ` technical   h      p 40      ` college   h      p 19      ` postgraduate   h      p 3 p     ` Total 0  0h(#(#0h(#h(#0(#(#0p(#(#87` p(#p(# ////  Othercharacteristicsoftheworkersenrolledinthestudywhichmaymakethestudymoredifficulttogeneralizeincludesthefollowing:      "0 `   71(83%)oftheworkershad5ormoreyearsexperienceinthechemicalindustry; ݌O ` (#` (# Ќ        "0 `   85(98%)ofthemhadparticipatedintraining;and ݌?` (#` (# Ќ        "0 `   theaveragetraininghoursspentonhealthandsafetywas71.82hoursayear. ݌/` (#` (# Ќ    Someadditionalconsiderationsidentifiedinthe//Lehto//studythatmaybeusefulinlabeldevelopmentarelistedbelowandtakendirectlyfromtheANSI//Z400.1//Standard:  1.0 ` Briefstatementsusingcommonvocabularywordsaremorelikelytobeunderstood;` (#` (#   2.0 ` Majoremphasisshouldfocusonwarningsforsevereandhighlikelihoodhazards;` (#` (#   3.0 ` Labelwarningsareforalertingratherthaneducating;` (#` (#   4.0 ` Prescriptivephrases(phrasesthatrecommendpositiveaction)areusuallymore  effectivethatproscriptivephrases(phrasesthatprohibitaction); ` (#` (#   5.0 ` Workerfeedbackonphrasescanbeusefulindevelopingnewstatements;andp` (#` (#   6.0 ` Theuseofcomplexconditionalsanddoublenegativesshouldbeavoided.`` (#` (#   Researchmustbedonetodevelopacommunicationtoolthatwillbe comprehensive without o @ sacrificing comprehension .AsgoodadocumentastheMSDSmaybewhenitispreparedusing _!0 ANSI//Z400.1//guidance,itstillmaynotbethebesttoolfortransferringinformationtoworkers.Therehavebeenconsistentobservationsthatcorroboratethisconcern.InanevaluationoftheoveralleffectivenessofMSDSsasacommunicationtoolforthehealtheffectsofhazardouschemicals,//Hadden//(1989)foundMSDSstobetoolongandladenwithtechnicaltermsunfamiliartomostworkers.////  Inthefinalanalysisoftheirstudyevaluatingthecontent,structureandformatofMSDSs,Cohenetal.(1989)suggestthatMSDS,bythemselves,areapoormeansofinformingworkersofhazardstowhichtheymaybeexposedforthefollowingreasons:1)muchofthetechnicalinformationhaslittlemeaningtotheaverageworkerandmayevenfrustratetheworkersabilitytoreadotherportionsoftheMSDSsthathaveinformationpertinenttohazardrecognitionandsafepractices;2)informationdepictinghazardousconditions,signsorsymptomsofexposure,andsafehandlingproceduresarewrittengenericallyandworkersmayhavedifficultyseeingtheconnection -`'* betweentheirownuseofthechemicalandtheinformationontheMSDS;and3)theinformationmaycontaintermstoodifficulttounderstandortoobriefandvaguetoactuallygeneratetheconcernthatworkersshouldhaveregardingsafeuseofthechemicals.  ThemostrecentstudyevaluatingthecomprehensibilityofMSDSandcomparingtheefficacyoftheInternationalChemicalSafetyCard,theANSI//Z400//format,andanOSHAtypeformwasdonewithapopulationofunionizedandnonunionworkersatalargenational(U.S.)researchlaboratory,Phillips,(1997).Theemployeesrepresentedalargenumberofcrafts,suchaspainters,//carpentars//,truckdrivers,andgenerallaborers.NinetyfivepercentoftheunionworkersstudiedweretrainedonMSDS.ThirtyninepercentoftheworkersfoundtheMSDS difficulttounderstand.OnethirdoftheworkersfoundtheMSDSformattobeconfusing.EachworkerwaspreandposttestedforknowledgeregardingahazardouschemicalbeforeandafterreadinganMSDS.Allthreeformatsimprovedtheworkersknowledgetoacertaindegree.Inrankorder,theInternationalChemicalSafetyCard(//ICSC//)fairedthebest,theOSHAtypeformsecond,andtheANSI//Z400//rankedlast.Variabilityexistedonspecificquestions.Forexample,      `  "0   The//ICSC//formwassignificantlybetterthattheOSHAtypeformat ? answeringquestionsonchronicandimmediatehealtheffects. ݌  (# (# Ќ     `     "0   Boththe//ICSC//andtheOSHAtypeformweresignificantlybetterthanthe  ANSI//Z400//inansweringquestionsonfirerelatedquestions. ݌  (# (# Ќ     `     _"0   TheOSHAtypeformwassignificantlybetteronspillresponsequestions. _z݌ (# (# Ќ        `  ?"0   TheANSI//Z400//wasnotsignificantlybetterinanyquestionaskedonthe  comprehensibilitytest. ?f݌  (# (# Ќ  ////  ThekeyandcorroboratingfindingofPhillipsstudywasthatathirdoftheinformationprovidedonanyoftheMSDSformatswasnotunderstood.Andyet,onceagainthisisabestcasescenariothestudypopulationwereliterate,trainedworkerswhospokeEnglishastheirfirstlanguage.////  ThefollowingisaselectionofrecommendationsmadebyPhillipsinlightofhisstudy:      `  "0   AneffortshouldbemadetounderstandwhyworkersfoundMSDSdifficult _!0 tounderstandandreadandwhyonethirdoftheinformationwasnot//transfered//. Q݌  (# (# Ќ        `  "0   ݀FutureMSDSresearchshouldincludemoreethnicallydiversepopulations. ݌/$  (# (# Ќ        `  l"0   0 (# (# Studiesshouldincludeinternationalpopulations. l݌%!(#(# Ќ  0  0` (#(#    "0 ` (#` (#  UntilfurtherquantitativeresearchtoevaluatetheefficacyofvariousMSDS &" formatsiscompletedandconsideringthefindingsofthislimitedresearch,cautionshouldbeusedinmovementtowardwidespreadadoptionofthenewANSI//Z400//structure.Effortsshouldbemadetfurtherfieldtestthenewformat. ݌  (# (# Ќ    Indesigningsafetycommunication,//Laughery//and//Brelsford//(1993)offeredthefollowing//principles://  1.0 ` Knowthereceiver.,p&)` (#` (#   2.0 ` Whenvariabilityexistsinthetargetaudience,designforthelowendextreme.-`'*` (#` (#   3.0 ` Whenthetargetaudienceconsistofsubgroupsthatdifferinrelevantcharacteristics, / consideremployingawarningsystemthatincludesdifferentcomponentsforthedifferentsubgroups. ` (#` (#   4.0 ` Markettestthewarning//system.` (#` (# //  Thecorollarytoprinciple#3isdonottrytoaccomplishtoomuchwithasinglewarning.Theauthorssuggested anexampleofwherethiscorollarymaybeviolatedisthecurrentOSHAguidelinesregardingthevarietyofsubgroupsinthetargetaudienceformaterialsafetydatasheets(MSDSs).Thesesubgroupsincludetoxicologists,safetyengineers,managers,physicians,andendusers(suchasthelaborerusingthestuff).Ifthewarningsystemdoesnotincludecommunicationmedia,inadditiontotheMSDS,itisprobablydestinedtofail(//Laughery//and//Brelsford//,1993).////  Importanttonote,isthefactthat//ICSC//arepreparedwiththeexplicitintentthatthey beusedattheshopfloorlevelbyworkers.The//ICSC//werefieldtestedintheinitialstagesofdevelopment.Withregardtomarkettestingthewarningsystem,principle#4,//Laughery//and//Brelsford//furthernotethat Despitethedesignersknowledgeofreceivercharacteristicsandeffortstoapplythatknowledge,warningsshouldbemarkettestedtoassesscomprehensionandbehavioralintentions.Theyalsoobservethat suchminimaleffortsareseldompartofthewarningdesignprocess. `   ////  Anotherstudynotingthechallengesthatmustbeovercome,examinedtheaccuracyofthedatapresentedwithintheMSDS(//Kolp//etal,1994).Thestudyevaluatedtheaccuracyandcompletenessofdataon150randomlyselectedMSDSinfivecategoriesofinformation:  1)chemicalidentificationofhazardousingredients,  2)reportedhealtheffects,  3)suggestedfirstaidprocedures,  4)recommendedpersonalprotectiveequipment(//PPE//),and  5)exposurelevelregulationsandguidelines.  Anexpertpanelreviewestablishedthatonly11%oftheMSDSswerefoundtobeaccurateinallofthefollowingfourareas:healtheffects,firstaid,//PPE//,andexposurelimits.Further,thehealtheffectsdataontheMSDSs frequentlyareincompleteandthechronicdataareoftenincorrectorlesscompletethantheacutedata.////  Thehumanhealtheffectsdataaregenerallythoughttobesparsegiventhenumberof hazardouschemicalsbeingusedinworkplaces.However,insomeinstancesithasbeennotedthatMSDSsareparticularlyfaulty.PaulandKurtz(1994)notethatthereproductivehealthhazardsweredeficientinalimitedstudyofMSDSswhentheylookedatMSDSsforleadandethyleneglycolethers,twosubstanceswithwelldocumentedriskstoreproductivehealth.   [   [24O  3  .  3  .  3  3  0    HAZARDCOMMUNICATIONASAWHOLEANDDYNAMICPROCESS Ԁ [݌+%((#(# Ќ    Theliteratureontheelementsofhazardcommunicationhavebeenreviewedabove.The -`'* relativemeritsofthesepartsastheycontributetothespecificgoalsofhazardcommunicationintheworkplacehasnotbeenevaluated.TheOfficeofCancerCommunicationattheU.S.NationalCancerInstitutepublishesMakingHealthCommunicationWork:PlannersGuide(NIH,1992)(hereafter  Guide)whichprovidesanexcellentframeworkforevaluatingoverallhazardcommunicationpolicies  andpractices,aswellashazardcommunicationprogramswithinworkplaces.Inaddition,thisframeworkcanbeusefulinmakingrecommendationsforthefuturedirectionofhazardcommunicationbothwithintheU.S.andinternationally.TheGuidedividesthehealth   communicationprocessintosixstages.Thesestagesareusedbelowtoelaboratequestionsthatmaybeusefulindirectingfutureresearchneedsandpolicyanalysis.ThequestionswhichareposedinX%XXXb   italics#XbXXX%##aretakendirectlyfromtheGuide.Theadditionalquestionsareposedbytheauthorsofthis p  report.   ` STAGE1Planningandstrategyselection 󀄀Thefirststageisthefoundationforany o@  communicationprogram.Acarefulassessmentoftheproblemmustbeoutlined.Regardinghazardcommunication,the problemisthatworkersneedmeaningfulinformationaboutthehazardouschemicalswithwhichtheyworkortowhichtheymaybeexposedduringthecourseoftheirworksothattheycanmakeinformeddecisionsabouttheirbehavior.Thefollowingquestionsshouldbeposedtoassistinplanning:      +"0 `   Whatisknownaboutthehazardouschemicals,dangers,andrelatedhealtheffects? +R݌` (#` (# Ќ        !"0 `   Whichexpertsknowaboutthespecifichazardouschemicalsandtheirassociated  effects? !<݌ ` (#` (# Ќ        "0 `   Whereisinformationresidingandhowcanitberetrieved? 7݌` (#` (# Ќ        "0 `   Howdoesinformationgetintothedatabasesthataretypicallysearchedwhenpeople  prepareMSDSs?  ݌ ` (#` (# Ќ        "0 `   Whatarethelimitationsoftheexistentdatasources? ݌` (#` (# Ќ        "0 `   Howcanweimprovethequantityandqualityofdataavailable,particularlywith p regardtohumanhealtheffects? ݌ ` (#` (# Ќ        "0 `   Forthosecompanieswhichdonothaveateamofhealthandsafetyprofessionalsto P assistinMSDSpreparation,howisaccessassuredtotheappropriateinformationsources? ݌ ` (#` (# Ќ        !"0 `   Shouldtherebeminimumrequirementsforthepersonresponsibleforthepreparation O"  oftheMSDSs?(Suchconcernsmaybecomeincreasinglyimportantintheinternationalarena.) !D!݌ ` (#` (# Ќ  X%XXXb0  Whatnewkindsofinformationwillbeneededtoplantheprogram?Whoisthetarget %! audience?Whatisknownaboutthem?#XbXXX%p"#ԀAveryimportantcharacteristicofthe target &" audienceforhazardcommunication,Americanworkers,isthatanestimated22%ofthemarefunctionallyilliterate.(Thiswillcertainlyvaryinternationally.) (#(#   X%XXXbOverall,whatchangeisplannedtosolveorlessentheproblem? ("%   Whatmeasurableobjectivescanbeestablishedtodefinesuccess#XbXXX%Y$#?Theonlymeasurable )#& 0  objectivesforsuccessthathavebeenappliedinmostworkplacesarewhetherornotcomponentpartsofthehazardcommunicationstandardrequirementsareinplace.Anotherobjective,atthe//worksite//level,maybetoimprovetheknowledgeandchangethebehaviorofmanagementandworkersastheyapplytohazardouschemicals.-`'*(#(# 0  X%XXXbHowcanprogressbemeasured?#XbXXX%&#ԀAtthe//worksite//level,measuresofprogressshouldbe / developed. (#(# X%XXXb0  Whatshouldthetargetaudiencebetold?#XbXXX%'#ԀAlthoughthereislittledebateastowhat  informationiscriticalforworkerstohave,therehasbeenmuchwrittenaboutthedisadvantagesofincludingcomplextechnicalinformationwhichthendiminishestheoverallgoalstoachievecomprehensionandultimatelybehavioralchanges.InplanningtheInternationalChemicalSafetyCards,aprocesswascreatedtoestablishwhatshouldandshouldnotbeplacedonthecard,thathaveworkersastheprimaryaudience.TheANSI//Z400.1//Standardcreatedadocumentformultipleaudiencesthat,inturn,maynotbethebesttooltocommunicateinformationexpresslytoworkers. (#(# // STAGE//2SelectingaMediumand//Materials //񀄀Inselectingthemediumandmaterialsto P  communicatethemessage,thefollowingquestionsshouldbeposed:0  Arethereanyexistingmaterialswhichcouldbeadaptedfortheprogram?Thisquestion _0  goestotheheartoftheinternationalharmonizationefforts.Thematerialsselectedshouldbebasedonsoundprinciplesofcommunicationandshouldbesubstantiallyevaluatedbeforethefinalselectionismade. (#(# 0  X%XXXbWhichmediaaremostappropriateforreachingthetargetaudience#XbXXX%X-#?Thefollowing  questionsshouldbeansweredregardingthemedium: (#(#    ` X%XXXbԄDoesitofferaccurate,complete,andrelevantmessages?    0 ` Isitappropriateforyourtargetaudienceinformat,style,andreadabilitylevel? ` (#` (#    ` Isitavailable?   ` Coulditbemodifiedtobecomeappropriate?#XbXXX%b.#  Anothersetofquestionsifyouareconsideringtheuseofexistingmaterial:  X%XXXb ` Howwerethemessagesdeveloped?     ` Werethematerialstested?   ` Howhavetheybeenused?   ` Weretheyeffective?#XbXXX%V0# P   (ThiswillbeanextremelyusefulsetofquestionsfortheinternationalharmonizationprocessandtheU.S.nationalprocessbetweenagencies).  X%XXXbԀWhatformatswillbestsuitthemediumandthemessages#XbXXX%/2#?Thescientificliterature O"  regardinghazardcommunicationformatsisfairlyprescriptiveandshouldbefollowedinreformulatingthethinkingregardingMSDSintheU.S.especiallyasinternationalharmonizationisconsidered.Itappearsevidentthatasimplerinformationsheetwithinformationspecifictoworkersimmediateneedscouldbeeffective.//Any//newformatshouldbeextensivelytestedwithavarietyofworkerpopulations̀ STAGE3DevelopingMaterialandPretesting Themessageshouldbeevaluatedandthetarget ("% audiencesreactionstothemessageshouldbeassessed.DuringthecourseoftheANSI//Z400.1//StandarddevelopmentastudywasdonetodeterminethecomprehensibilityofselectedhazardcommunicationphrasestobeusedonX%XXXblabels.#XbXXX%5#ԀAlthoughthiswasausefulstudy,the +%( comprehensibilityofanX%XXXbMSDS#XbXXX%6#initsentiretywasnotevaluated. ,p&)   ThereisasetoffourquestionsthattheGuiderecommendspertainingtotheaudience: -`'* ?> 2 ?<G@/< ?    DOESTHEAUDIENCE:   ` X%XXXbԄUnderstandthemessage?     ` Recallit?   ` Acceptitasimportant?   ` Agreewiththevalueofthesolution(safebehavior)?#XbXXX%=8#   ////  Itiscriticalthattheanswerstothesequestionberigorouslyinvestigated.Howtheaudiencerespondstothemessageformatisalsoimportant.WhenRohmandHaas,thechemicalcompany,wasdevelopingtheirowninhouseMSDS,theydidastudytodeterminetheiremployeespreferencefortheformatoftheMSDSandpreferenceforthetypefont.Theyweresurprisedtodiscoverthattheemployeeshadanoverwhelmingpreferenceforoneofthethreeformatsthatwerepresentedandoneofthefonttypesovertheothertwo.(//Ignatowski//,1997)RohmandHaasthenproceededtodeveloptheirinhouseMSDSbasedontheemployeepreference. STAGE4Implementation 󀄀Theimplementationphasemustincludethecoordinationofallthe O  stakeholdersintheprocesssothattheoverallprogramgoalscanbeachieved.Programcomponentsshouldbeperiodicallyreviewedandrevised,ifnecessary.Duringthisphase,itisimportanttoknowwhetherthemessageismakingitthroughtheintendedchannelsofcommunication.Whenhazardcommunicationprogramsarebeingimplemented,akeyplayerintheprocessisthelinesupervisor.Inareviewofhazardcontrolprogramswithexemplarysafetyandhealthperformancerecords,itwasfoundthatthetimedevotedtotrainingandtheinvolvementofsupervisorswereimportantfactors(Cohenand//Colligan//,1995).Thesupervisorsmustunderstandthelinktheyareresponsibleforinthechainofhazardcommunication.////  X%XXXbIsthetargetaudiencepayingattentionandreacting?#XbXXX%?#Theanswertothisquestionislargely  dependentonthequalityofthetrainingthatisofferedinsupportoftheothercommunicationtools,suchaslabelsandMSDSs.Theuseofvideotraining,intheabsenceofanyothercommunicationprocesswillnoteffectivelycapturetheaudiencesattention,especiallyifthevideotrainingisoffered,asitoftenis,attheendofaworkersshift.// STAGE//5AssessingEffectiveness Theprogramshouldbeassessedbyanalyzingtheresultsof O"  measurementsestablishedintheplanningstage.Weretheprogramsobjectivesmet?Mostemployerswilltranslatethisintodeterminingwhetherornotthevariouscomplianceactivitieswereimplemented.Theliteraturereportingthatknowledgeorimprovedworkpracticesresultedfromhazardcommunicationprogramsislimited.Thisislargelybecauseitisnotrequiredtoevaluatetheeffectivenessofaprogram(andbecauseresearchfundingisvirtuallynonexistentforsuchendeavorstobedonebyuniversityandresearchinstitutionsandotherswhomightpublishtheresults).Thesetypeofevaluationsmaybetakingplace inhousewithinindustries,howeverreportsoftheresultsofsuchevaluationsarenotgettingpublished.////  IntheU.S.autoindustry,hazardcommunicationhasbeentakenveryseriouslyandasubstantialinvestmenthasbeenmadetoimplementtheirprograms,whichwerecreatedandimplementedbythecombinedparticipationoftheindustry,theUnitedAutoWorkers,andthe -`'* UniversityofMichigan.InanevaluationoftheFordMotorCompany,hazardcommunicationtrainingprogram,itwasdeterminedthat moreinteractive,trainerintensivedeliverymethodstosmallergroupswereassociatedwithmorepositiveeffectsonreportedtrainingusefulnessandchangesinworkpracticesandworkingconditions(Robinsetal.1990).Workmustcontinuetobesupportedtoidentifythecriticalvariablesthatwillpredicttrainingsuccesses. STAGE6FeedbacktoRefineProgram 󀀄Insomeways,Canada,theU.S.,theEuropean   Communityandotherswhohavedevelopedsomehazardcommunicationmechanismsarenowinstage6,atimeofevaluatingfeedbackandrefiningtheirprograms.Thisis,onceagain,especiallytrueasharmonizationisconsidered.Themoreinformationthatcanbegathered,themorelikelyitisthatthefollowingquestionswillbeanswered:   ` X%XXXbԄWhyhastheprogramworkedornotworked? P  0  0` (#(#Arethereprogramchangesthatshouldbemadetoincreasethelikelihoodofsuccessortoaddresschangesintheaudience,orproblemorsituations? ` (#` (#    ` Aretherelessonslearnedthatcouldmakefutureprogramsmoresuccessful? #XbXXX%/J# O  //  // [   [L24O  3  .  3  .  4  3  0    ADDITIONALRESEARCHNEEDS L [LIL݌/(#(# Ќ    Basedonthereviewoftheliteratureonthemajoraspectsofhazardcommunication,itappearsthattherearetwosignificantresearchneeds.ThefirstisunderstandinghowwecansignificantlyimproveMSDSsotheycanbecomeamoremeaningfultoolforcommunication.Andthesecondishowandwherethe communicationinhazardcommunicationoccurs.Whatarethevariablesthatwillpredictthataworkerwillreceiveandunderstandtheinformationnecessarytomakeinformeddecisionsinhis/herworkplace?Whatarethevariablesdeterminingsafeandinformedbehaviorsregardinghazardouschemicals?////  RegardingthequalityofMSDS,wemayneedtoconsiderwhethertheMSDSinitscurrentformisthebestwrittencommunicationtooland,ifnot,lookseriouslyforhowitmightbealtered.Regardlessofwhatultimateformittakes,itshouldbeinastandardformat.Thereshouldbesomeminimumqualificationforthepersonpreparingthedocument..Thereshouldbeasystemofrandom,periodicreviewofaselectionofMSDSbyathirdpartytodeterminetheaccuracyofthedatarepresentedonthesheetsandsanctionsappliedfornoncompliance.//  //  3    3#R24O  4  3  0    CONCLUSIONS  R 3#RNR݌%!(#(# Ќ  //  Aswemoveintothe21stCentury,itisimportanttoconsiderwhoandwhataremovingwithus.Thereareimmensechangesininformationtechnologythathaveeruptedinthelastdecadeforthetransmissionofinformation,//factsheets//,checklists,trainingcurriculumandotherofthebasictoolsofcommunication.Theinternationalpublichealthandsafetycommunitycouldbeconveningavarietyofgroupsofexpertstohelpformthevisionandstrategicplanforbringinghighquality,effectivehazardcommunicationprogramsintothenextcentury.Forexample,anexpertpanelincomputerandinformationtransfertechnologycouldbeconvenedtodiscusshowtodeliveraglobalhazardcommunicationprogram.Suchinformationcanhelptoinformthelongertermgoalsthat -`'* shouldbeconsideredinthenextphaseofinternationalhazardcommunicationdevelopmentandharmonization.  Nationalcampaignshavehelpedtochangeattitudesandbehaviorsregardingseatbelts,helmets,andrecycling.Socialmarketinghasproventobeapowerfulforceforchangingbehavioronagrandscale.Itmightbehelpfultoelicitthesupportofexpertsinsocialmarketingtoheartheirideasforthebestmessagestoachievehazardcommunicationgoals.////  Mostimportantly,itmustbestressedthatsoundscienceandprovenefficacyprovidethefoundationforournextstagesintheveryimportantmissionofachievingworkerandcommunityhealthandsafety.̀  o@      3   /XbXXXb 3IY24O  5  3  0    BIBLIOGRAPHY