... whenyouanswer,includeanypartnersthatyoumayhavealreadytoldyourintervieweraboutandanythatyoudidnotwanttotalkwiththeinterviewerabout.’ The self-administeredversionsofthe‘‘numberofpartners’’dataappearintables28,30,and32.Numberofpregnanciesandnumberoflivebirths—ThePREGNUMrecode(usedintable3)specifiesthetotalnumberofpregnancies,includingacurrentpregnancy,reportedbythewoman.ThePARITYrecode(usedasthedependentvariableintable4and5)givesthetotalnumberofbabiesborn,accountingformultiplebirths,fromallpregnanciesreportedbythewomanintheinterviewer-administeredportionofthesurvey.Intheself-administered(AudioCASI)partoftheinterview,womenwereaskedagainaboutpregnanciesendinginabortion.Takingoneofseveralpossibleapproaches,afewrecodesweredefinedtocombinePage110[Series23,No.19MethodsInterviews ... 109twoeventsiscalculatedbysubtractingthedateoffirstintercoursesincemenarchefromthedateoffirstmarriage.Iffirstintercourseaftermenarcheoccurredafterfirstmarriage,orbotheventsoccurredinthesamemonth,itisindicatedaccordinglyintable26.Monthspregnantwhenprenatalcarebegan—Forallpregnanciesendinginanywayotherthaninducedabortionin1991–95,womenwereaskedtoreportiftheyhadevervisitedadoctor,midwife,orclinicforprenatalcare.Thosewhoreportedanyprenatalcarewerethenaskedhowmanyweeksormonthspregnanttheywerewhentheyfirstwentforprenatalcare.ThePNCAREWKrecodeindicatesthenumberofweekspregnantatfirstprenatalvisit.Table83givesthepercentdistributionofPNCAREWKforallpregnanciesthatresultedinlivebirthin1991–95 .The categoriesshownare‘‘lessthan3months,’’‘‘3–4months,’’and‘‘5monthsormoreornoprenatalcare.’’‘‘Lessthan3months’’isdefinedaslessthan13weeks,tobeconsistentwiththedefinitionofearly(firsttrimester)prenatalcareinpreviousreportsfromtheNSFG.Mother’seducation(recode=EDUCMOM)—‘‘Mother’seducation’’wasclassifiedaccordingtothehighestgradeofregularschoolorcollegecompletedbytherespondent’smotherormother-figure.ThismeasurediffersslightlyfrompastNSFGreportsthatwerebasedonaquestionaskingforeducationof‘‘mother’’or‘‘stepmother.’’TheCycle5versiondiffersbecausetheidentityofthemotherormother-figurewasascertainedinamorepreciseway:ifthechildhoodlivingarrangementsshowedthattherespondentlivedwithhernaturalmotheratsomepointandneverlivedwithafemaleparentotherthanthenaturalmother,thenthenaturalmotheristhepersonaboutwhomeducationwasasked.Otherwise ,the respondentwasasked:‘‘Who,ifanyone,doyouthinkofasthewomanwhomostlyraisedyouwhenyouwereateenager?’’Thepersonreportedbecomesthepersonaboutwhomeducationisasked.Respondentseligibleforthatquestionwereallowedtorespond‘‘nosuchperson,’’and0.7percentofsamplecasesfellintothiscategory.Thiscategoryislabeled‘‘nomother-figureidentified’’inthetables.Anadditional4.3percentofinterviewedwomenhadmissingdataonmother’seducation,andthesevaluesfortheEDUCMOMrecodewereimputed.Nonvoluntaryfirstsexualintercourse—Thevoluntary/nonvoluntarystatusoffirstintercoursewasdeterminedfromtwoquestionsintheinterviewer-administeredinterview.First,respondentswereasked:‘‘LookingatthescaleonCardC-3,whichnumberwouldyousaycomesclosesttodescribinghowmuchyouwantedthatfirstsexualintercoursetohappen?Onthisscale,a1meansthatyoureallydidn’twantittohappenatthetime,anda10meansthatyoureallywantedittohappen.’’Womenwhorespondedtothisquestionthattheirfirstintercoursewasarapeareclassifiedashavingnonvoluntaryfirstintercourse.Respondentswhoreportedanythingotherthan‘‘rape’’wereasked:‘‘Wouldyousaythenthatthisfirstsexualintercoursewasvoluntaryornotvoluntary?’’Respondentsanswering‘‘notvoluntary’’arealsoclassifiedashavingnonvoluntaryfirstintercourseforthetable(table21).Numberofhusbandsorcohabitingpartners—TherecodeFMARNOgivesthenumberoftimesawomanhasbeenlegallymarried.Womenwerealsoaskedthenumberofothercohabitingpartnerstheymayhavehad.Table37isbasedonthetotalnumberofhusbandsandcohabitingpartners,computedasfollows:FMARNO+(1ifcurrentlycohabiting)+(#ofothercohabitingpartners)Husbandswithwhomawomanalsocohabited(outsideofmarriage)arecountedonlyonce,ashusbands.Numberofmalesexualpartners—Eachwomanwhohadeverhadvoluntarysexualintercousewasaskedtoreportthenumberorrange(low-high)ofmaleswithwhomshehashadsexual(vaginal)intercourseinherlife,asofthedateofinterview.ShewasalsoaskedtoreporthernumberofmalesexualpartnerssinceJanuary1991(roughlythelast5years)andinthelast12months.Ifarangeofpartnerswasreported ,the numberofpartnerswascalculatedastheaverageofthelowandhighnumbers .In thosefewcaseswheretherespondentreportedzeropartnersforthelowendoftherangeandonepartnerforthehighend,thenumberofpartnerswassetequaltoone.Respondentswhohadneverhadintercourseorwhohadonlyhad(first)nonvoluntaryintercoursewereassignedzero.Thereweretwoversionsofthisseriesofthreequestionsaboutnumberofmalesexualpartners.Thedescriptionaboveappliestotheversionintheinterviewer-administeredportionofthe survey, andtheseresultsarepresentedintables27,29,and31.ThesequestionswererepeatedintheAudioCASI,orself-administered,portionofthesurvey,attheendoftheinterview.Eachpairofquestions(interviewer-administeredandself-administered)waswordedalmostidentically,althoughtheAudioCASIquestionsaskedforaspecificnumberanddidnotallowarangetobegiven.Also,thequestionsintheself-administeredportionwereprefacedbythefollowinginstruction:‘‘ ... number of timeswill be lower in the survey data thanthey actually are in the population. The data in table 3 are from the interviewer-administered part of the NSFG.Including data from the self-administered...