6.How-to-Write-an-Abstract

1TheAbstractI.Generalfunctionsofanabstract(themostimportantcomponentofaprofessionalpaper,theveryfirstpartofaprofessionalpaperandaselfcontainedentity)1.miniaturizingthetext(acondensedstatementofthecontentsofapaper;ashort,conciseandhighlygeneralizedtext,amini-versionofthemainbody)2.decidingYesorNo(ForbusyreaderstheAbstract,sometimescalledtheSummary,maybetheonlypartofthepapertheyread,unlessitsucceedsinconvincingthemtotakethetimetoreadthewholepaper.Itmaydirectlyinfluencethepaper’sacceptancetoalearnedjournal.Agoodabstractshouldattractaudienceandconvincethemtoreadthewholepaper,andpromotethoroughdiscussionontherelevantsubject)3.expandingthecirculation(Becauseon-linesearchdatabasestypicallycontainonlyabstracts,itisvitaltowriteacompletebutconcisedescriptionofyourworktoenticepotentialreadersintoobtainingacopyofthefullpaper.AbstractingservicesmayusethetextofthetitleplustheAbstractandkeywordsfortheirsearchabledatabases.Forreadersindevelopingcountrieswithlimitedaccesstotheliterature,theabstractmaybetheonlyinformationonyourworkthatisavailabletothem.)II.linguisticfeatures1.limitedlength(200wordsisasensiblemaximumforarelativelylongerpaper,nevermorethan500words;50-100wordsmaysufficeforashortarticle.Asageneralrule,anabstractwillbeabout3-5%ofthelengthofthepaper,seldomlongerthan2/3sofapage.)2.categoriesofabstracts(Threekinds:descriptive,informationalandinformational-indicative.Adescriptiveorindicativeabstractusuallystatesthegeneralsubjectmatterofthedocumentthatfollows.Ittellsinaqualitativewaywhatthepaperorreportcontains.Thesetellreaderswhatinformationthedissertationcontains,andincludethepurpose,methods,andscopeofthereport,article,orpaper.Adescriptiveabstractwillnotprovideresults,conclusions,orrecommendations,andisusuallyshorterthananinformativeabstract-usuallyunder100words.Itspurposeistomerelyintroducethesubjecttoreader,whomustthenreadthedissertationtofindoutyourresults,conclusions,orrecommendations.2Aninformationalorinformativeabstracthighlightsthefindingsandresults,brieflyandquantitatively.Itisacondensedversionoftheresearchwork,withoutdiscussionorinterpretation.Itshouldbespecificandquantitative,givingonlyessentialdata.Thesecommunicatespecificinformationfromthedissertation,includingthepurpose,methods,andscopeofthereport,article,orpaper.Theyprovidethedissertationresults,conclusions,andrecommendations.Theyareshortbutnotasshortasadescriptiveabstract-usuallyanythingfromaparagraphtoapageortwo,dependinguponthelengthoftheoriginalworkbeingabstracted.Inanycase,informativeabstractsmakeup10%orlessofthelengthoftheoriginalpiece.Theinformativeabstractallowsyourreadertodecidewhethertheywanttoreadthedissertation.Theinformational-indicativeabstractisacombinedformthatbearsspecificinformationaboutprinciplefindingsandresultsandgeneralinformationabouttherestofthedocument.Thistypeifabstractoffersfewerdetails,instead,givingemphasistotheauthor’schiefcontribution.Indoingso,thefollowingabstracthighlightsthesametechnicalinformation.)3.completecontent(Theabstractisself-contained,unifiedandcoherentincontent.Aninformationalabstractshouldatleastcontainthefollowingthreeelements:astatementoftheproblem;astatementoftheapproachtosolvingtheproblem;theprincipalresult.)Abstractshouldinclude:Background:introducethetopic(optional)Purposeofthestudy:studyaimMethodsofthestudy:overviewofthemethodsFinding/Resultsofthestudy:overviewofresultConclusion,implicationofthestudyMotivation:Whydowecareabouttheproblemandtheresults?Iftheproblemisn'tobviouslyinterestingitmightbebettertoputmotivationfirst;butifyourworkisincrementalprogressonaproblemthatiswidelyrecognizedasimportant,thenitisprobablybettertoputtheproblemstatementfirsttoindicatewhichpieceofthelargerproblemyouarebreakingofftoworkon.Thissectionshouldincludetheimportanceofyourwork,thedifficultyofthearea,andtheimpactitmighthaveifsuccessful.Problemstatement:Whatproblemareyoutryingtosolve?Whatisthescopeofyourwork(ageneralizedapproach,orforaspecificsituation)?Becarefulnottousetoomuchjargon.Insomecasesitisappropriatetoputtheproblemstatement3beforethemotivation,butusuallythisonlyworksifmostreadersalreadyunderstandwhytheproblemisimportant.Approach:Howdidyougoaboutsolvingormakingprogressontheproblem?Didyouusesimulation,analyticmodels,prototypeconstruction,oranalysisoffielddataforanactualproduct?Whatwastheextentofyourwork(didyoulookatoneapplicationprogramorahundredprogramsintwentydifferentprogramminglanguages?)Whatimportantvariablesdidyoucontrol,ignore,ormeasure?Results:What'stheanswer?Specifically,mostgoodcomputerarchitecturepapersconcludethatsomethingissomanypercentfaster,cheaper,smaller,orotherwisebetterthansomethingelse.Puttheresultthere,innumbers.Avoidvague,hand-wavingresultssuchasvery,small,orsignificant.Ifyoumustbevague,youareonlygivenlicensetodosowhenyoucantalkaboutorders-of-magnitudeimprovement.Thereisatensionhereinthatyoushouldnotprovidenumbersthatcanbeeasilymisinterpreted,butontheotherhandyoudon'thaveroomforallthecaveats.Conclusions:Whataretheimplicationsofyouranswer?Isitgoingtochangetheworld(unlikely),beasignificant