15. Is drinking coffee good or bad for you? Is decaf better than caffeinated coffee? Mostpreviousstudiesthathavetriedtoanswersuchquestionshavebeenbasedonsurveysaskingpeopletoreporttheircoffeedrinking. In2005,Dr.RobertSuperkopresentedtheresultsofwhatis probablythefirsttrulyexperimentalstudyoftheissue. Subjectswere187volunteers. Theyweredividedintothreegroups. For8weekstheydrank: caffeinatedcoffee(black,3to6cupsper day);decaf(black,3to6cups); nocoffeeatall. Thecoffeedrinkerswereprovidedwithcoffeemachinesandpre-measuredservingsofgroundcoffee. Theresultsofthebloodtestsonsubjects inthisfederallyfundedstudyweremixed. Mostwidelyreportedwasthefactthatforsomesubjects,decaffeinatedcoffeelowered“goodcholesterol”(HDL)significantly,by30%. Forsubjectswhowereoverweight,however,theeffectswerejusttheopposite: forthem,decafboostedHDLlevelsbyabout50%.
a.
Doesthefactthatparticipantswerevolunteersraiseproblemsofbias? How,ifatall,doesvolunteerbiasdifferinexperimentalresearchwheresubjectsarerandomlyassignedascomparedtosurveyresearchwherevolunteerswouldnot,bydefinition,bearandomsample?
b.
Whatothervariables,ifany,mightexplaindifferencesintheeffectsofdifferenttypesofcoffee?
c.
Wasthesamplebigenough? Howcanonestudymakedowith187subjectswhenanother—thepoliostudyinthepreviousexample—required400,000?