搜索
WeitzlabGuidetoGoodPaperWritingLastupdated5/1/2012Allscientificpapersshouldbewellwritten;thiswillmakethereaderwanttoreadthepaper,andwillincreasethelikelihoodthatthereaderwillactuallypayattentiontowhatyouaretryingtosay.Youmayhavebeentaughttowritelabreports,andhencepapers,asfollowingtheorderofintroduction,summaryofpreviousresults,experiment,resultsanddiscussion.However,inalmostallcases,ifyouwriteapaperwiththisstructure,itwillbejustlikealabreport:Boring!Thesearesuggestionsforwritingbetterscientificpapers.However,pleaserememberthat,likeallcaseswiththeEnglishlanguage,thesearenotrules,butareguidelines,andtherewillcertainlybetimesthattheydonotapplyorthattheyshouldbechanged.Nevertheless,dothisonlywithcareandforeknowledge.Conclusion:•Startwiththeconclusionfirst.Thisdoesnotmeantheconclusionsection,buttheconclusionyouhavearrivedat.Makeitonesentence,twoatmost.Haveasingle,keypointyouaretryingtomake.Thesearethemostimportantsentencesthatyouwillwrite,asthiswilldetermineexactlywhatthepaperisabout.Ifyouwritethisfirst,thenyoucanwritetherestofthepapertojustmakethispoint.•Ideally,youknowexactlywhatthemainconclusionisbeforeyouwritethepaper.Thesearetheeasiestpaperstowrite.However,therearetimeswhentheactualwritingofthepaperwillhelprefinewhatitisabout.Thusneverbeafraidtochangethesesentencesifyoulearnmoreuponwritingthepaper.Thiscanmeanrewritingpartsofthepaper,butthiswillinvariablyleadtoabetterpaper,somakesuretodothis.Mainpointofthepaper:•Ashortpaper(Science,Nature,PRLorotherlettersjournal)canonlymakeonemainpoint,andperhapshalfofasecondpoint.•Alongerpapercanmakeanadditionalpoint.However,rarelycanasinglepapermakemorethanacoupleofpoints.Thereshouldalwaysbeacentralpointforeachpaper.Introduction:•Theintroductionisthemostimportantpartofthepaper.Itiswhatwillmakethereaderwanttoreadmoreofthepaper.Itisoftenalsothemainparagraphthatdetermineswhetherornotapaperisaccepted.•Writetheintroductionafteryouhavewrittenthetwosentencesummary.•Typically,youwillwritetheintroductionwhenyoufirstbeginthepaper.Youwillthencompletelyrewriteitwhenyouhavefinishedthepaper,andyoureallyunderstandwhatthemainpointofthepaperis.•Followthisgeneralordertowriteaneffectiveintroduction:oHaveageneral,introductorysentenceaboutthetopic.2oInthefirstfewsentences,establishthatthetopicofthepaper,assummarizedinyourone-sentenceconclusion,isinterestingandimportant,preferablybothscientificallyandtechnologically.oAlotmayalreadybeknown,soreviewtheliterature,butveryconcisely.oConcludefromthisreviewthatthereisonekeythingthatisstillunknown,withoutwhichwecannotprogress.oThiskeythingisexactlywhatyourconclusionsentenceisabout;thismeansthatyouhaveintroducedexactlywhatyourpaperisabout.oFinishtheintroductoryparagraphwithastatementabouthowifweunderstoodorknewtheinformationaboutthetopicofthepaperitwouldbeveryimportant.oThenstartanewparagraphwhichveryconciselydescribeswhatyouhavedoneandwhatyouhavelearned.Thisshouldbeashortparagraph,andshouldessentiallysummarizethepointmadeintheconclusionsentenceyouhaveprepared.Figures:•Choosethefigures,andwritecaptionsbeforeyouwritethemaintextofthepaper.Theseshouldbechosentomakethepointthatisyourconclusion.Havingthefigureschosenfirstwillmakewritingeasier.•Discussthefiguresinthetext.Guidethereaderthroughthefiguresinthetextofthepaper.Thereadershouldnothavetostopandreadthecaptiontounderstandwhatisinthefigureandwhyitisimportant.Notethatthisisnotauniversalrule:bothScienceandNatureprefertohaveaverydetailedcaption,withverylittlediscussionofthefigureinthetext.•Whendescribingafigure,itisimportantthatthereaderknowwhattolookforbeforelookingatthefigure.Therefore,itisalwaysbesttodescribewhatyouareplottingorshowing,andthekeypointthatyouwantthereadertocomeawaywith,beforeyouhavethereaderlookatthefigure.Thereadershouldlookatthefigurewhenyouwrite‘inFig.x.’Thereader’seyeswillgothefigurethen,breakingtheflowofthetext,soyoushouldhavedescribedthefigurebeforeyousaythis.Thus,itisalwaysmorereadabletoputaphrase‘asshowninFig.x’attheendofasentenceafterthemainpointofthefigureisdescribedratherthanstartingasentencewith‘InFig.x,weshow.’•Thediscussionofafigureshouldappearwhereitisnatural.Forexample,ifthereareseveralpiecesofdatathatarecomparedinthefigure,eachsetcanbediscussedinseparateplacesinthepaper.Outline:•Youaretryingtomakethepointoftheconclusion,andyouaretryingtoprovidethemostconvincingcasetothereader.Thiswilloftenmeanpresentingthingsinanorderthatisdifferentthanthewayyoudiscoveredtheconclusion;don’tbehistorical,beasclearaspossible.•Itisusuallybesttowriteanoutlineafterwritingthesummarysentenceandchoosingthefigures.Thiswillmakewritingthepapereasier,andwillensurethatthewholepaperisstructuredtomakethepointinthesummarysentences.3Readability:•Allpapersshouldtellastoryandshouldbeinterestingforthereadertoread;thiswillhelpensurethattheyareread,understood,andthereforepeoplepayattentiontothem,whichisthepurposeofwritingapaper.•LongpaperscansometimesfollowthetraditionalformatofIntroduction,Experiment,Results,DiscussionandConclusion.However,theyoftenwillbeclearerbynotfollowingthisformat.Theformatchosenshouldma