6.5 Application of combined consumerbehaviour ± food
safetystudies
Food safetyconsumerstudies oftenfocuson measuresof self-reported behaviour
or attitudestowards foodsafety ratherthanactual observationsof consumer
behaviour and whatthis mightimplyfor the incidence of foodpoisoning. Clayton
and colleagueshaveattempted to validateself-reported behaviours by comparing
these data withobservationaldata (Claytonet al., 2002).The resultsindicate that
some important actionssuchas hand-washingweremorefrequentlyreported
thanactuallyenactedby respondents.Comparisonof observational datawith
safety protocolssuchas HazardAnalysisand CriticalControl Points (HACCP)
did, in fact,indicatethat consumerbehaviours wereverifiableagainstmicrobial
contamination (Griffith and Worsfold, 1994,Worsfold and Griffith, 1995;
Griffithet al., 1998).However, few researchstudies reported in two recent
reviewpaperson consumerbehaviourand relationto foodsafety, studied con-
sumer attitudesand risk perceptionas well as consumerbehaviourobservationsat
the samesample(Redmondand Griffith, 2003a,b). To our knowledge, the
relevant cognitive representationsof consumers, resultingconsumerbehaviours
and microbial contaminationhavenot beenstudiedsimultaneously.Thereis
some convergence of resultsacross differentstudies.However, a fundamental
understanding of what consumer behaviours and activities resultin whatlevelsof
microbial contamination, how thesebehaviours and activities varyamong
individuals,and the role of humaninformationprocessingand affect(i.e. emotion
suchas anxiety or fear)in developing effective communicationstrategies,
remains largely unexplored. It is suggested thatthe only effective wayto
understandthe relationshipbetweenthesedifferentareasis therefore to integrate
social and naturalsciences,whichmayindicatethe needfor a newresearch
agendain this area.
6.6 The needfor moreintensive cooperationbetweennatural
and socialscientists
To be ableto tailorinformation campaigns to individualinformation needs,
much more detailedinformation on risk-related attitudesand behaviours is
needed, as wellas what the consequences of theseare for individualhealth
outcomes.In the caseof a national campaign targetedat population level
audiences,average risk levels to consumersare generallyapplied, and it is
unlikely that individualconsumers (possiblythose most at risk)will attendto the
information contained in risk messages. In contrast, tailored or targeted
campaignsmustfocuson the informationneedsof groupsor segmentsin the
population.To be ableto designa successful campaign, realisticestimates of
risksshouldbe communicated, along withany uncertaintiesaboutthese risk
estimates(if theyexist). Failureto do so mayhavea negative impacton trustin
the informationsource(Freweret al., 1996).Thuswhen targetingdistinct groups
the relation betweenspecificbehaviourand specificrisks shouldbe known.
112 Handbookof hygiene controlin the foodindustry