chapter  2
9 Pages

Taking a sample

Whengatheringorselectingdataforanalysis,achoicehastobemade betweentotalenumeration,i.e.countingormeasuringalltheoccurrences oftheparticularphenomenon,andsampling,i.e.countingormeasuringa selectedsub-setoftheseoccurrences.Intheformercase,ouranalyseswill provideanabsolutestatementonthistotalityofdata,and-withinthe accuracyoftheobservations-ourstatementwillbe'correct'.Ifthetotal enumerationisreallyofalltheoccurrencesthatexist,thenthisissatisfactory.If,ontheotherhand,wehavetotallyenumeratedalltheemployedpopulationinjustonetown(recordingtheiroccupations,income, etc.),thenourresultsarerelevanttothattownalone.Theydonotprovide uswithresultsfromwhichwecangeneralizeaboutconditionsin,for example,allothertownsofthesamesizeinthesamecountry.Yetitis thisverygeneralizationwhichisnecessaryifgeographersaretoattemptto understandandexplainrelationshipsintherealworld,andwhichhas beencharacteristicofdevelopmentsinthesubjectsincethe1960s.