Therelationships(bothconceptualandmathematical)betweenBayesiananalysisandstatisticaldecisiontheoryaresostrongthatitissomewhatunnaturaltolearnonewithouttheother.Nevertheless,majorportionsofeachhavedevelopedseparately.OntheBayesianside,thereisanextensivelydevelopedBayesiantheoryofstatisticalinference(bothsubjectiveandobjectiveversions).Thistheoryrecognizestheimportanceofviewingstatisticalanalysisconditionally(i.e.,treatingobserveddataasknownratherthanunknown),evenwhennolossfunctionistobeincorporatedintotheanalysis.Thereisalsoawell-developed(frequentist)decisiontheory,whichavoidsformalutilizationofpriordistributionsandseekstoprovideafoundationforfrequentiststatisticaltheory.AlthoughthecentralthreadofthebookwillbeBayesiandecisiontheory,bothBayesianinferenceandnon-Bayesiandecisiontheorywillbeextensivelydiscussed.Indeed,thebookiswrittensoastoallow,say,theteachingofacourseoneithersubjectseparately.