HowtoReadaScientificPaper如何阅读科学性论文.docx

HowtoReadaScientificPaper如何阅读科学性论文.docx

《HowtoReadaScientificPaper如何阅读科学性论文.docx》由会员分享，可在线阅读，更多相关《HowtoReadaScientificPaper如何阅读科学性论文.docx（8页珍藏版）》请在冰豆网上搜索。

HowtoReadaScientificPaper如何阅读科学性论文

HowtoReadaScientificPaper

Themainpurposeofascientificpaperistoreportnewresults,usuallyexperimental,andtorelatetheseresultstopreviousknowledgeinthefield.Papersareoneofthemostimportantwaysthatwecommunicatewithoneanother.

Inunderstandinghowtoreadapaper,weneedtostartatthebeginningwithafewpreliminaries.Wethenaddressthemainquestionsthatwillenableyoutounderstandandevaluatethepaper.

1. Howarepapersorganized?

2. HowdoIpreparetoreadapaper,particularlyinanareanotsofamiliartome?

3. WhatdifficultiescanIexpect?

4. HowdoIunderstandandevaluate thecontentsofthepaper?

1.Organizationofapaper

Inmostscientificjournals,scientificpapersfollowastandardformat.Theyaredividedintoseveralsections,andeachsectionservesaspecificpurposeinthepaper.Wefirstdescribethestandardformat,thensomevariationsonthatformat.

Apaperbeginswithashort Summary or Abstract.Generally,itgivesabriefbackgroundtothetopic;describesconciselythemajorfindingsofthepaper;andrelatesthesefindingstothefieldofstudy.Aswillbeseen,thislogicalorderisalsothatofthepaperasawhole.

Thenextsectionofthepaperisthe Introduction.Inmanyjournalsthissectionisnotgivenatitle.Asitsnameimplies,thissectionpresentsthebackgroundknowledgenecessaryforthereadertounderstandwhythefindingsofthepaperareanadvanceontheknowledgeinthefield.Typically,theIntroductiondescribesfirsttheacceptedstateofknowledgeinaspecializedfield;thenitfocusesmorespecificallyonaparticularaspect,usuallydescribingafindingorsetoffindingsthatleddirectlytotheworkdescribedinthepaper.Iftheauthorsaretestingahypothesis,thesourceofthathypothesisisspelledout,findingsaregivenwithwhichitisconsistent,andoneormorepredictionsaregiven.Inmanypapers,oneorseveralmajorconclusionsofthepaperarepresentedattheendofthissection,sothatthereaderknowsthemajoranswerstothequestionsjustposed.Papersmoredescriptiveorcomparativeinnaturemaybeginwithanintroductiontoanareawhichintereststheauthors,ortheneedforabroaderdatabase.

Thenextsectionofmostpapersisthe MaterialsandMethods.Insomejournalsthissectionisthelastone.Itspurposeistodescribethematerialsusedintheexperimentsandthemethodsbywhichtheexperimentswerecarriedout.Inprinciple,thisdescriptionshouldbedetailedenoughtoallowotherresearcherstoreplicatethework.Inpractice,thesedescriptionsareoftenhighlycompressed,andtheyoftenreferbacktopreviouspapersbytheauthors.

Thethirdsectionisusually Results.Thissectiondescribestheexperimentsandthereasonstheyweredone.Generally,thelogicoftheResultssectionfollowsdirectlyfromthatoftheIntroduction.Thatis,theIntroductionposesthequestionsaddressedintheearlypartofResults.Beyondthispoint,theorganizationofResultsdiffersfromonepapertoanother.Insomepapers,theresultsarepresentedwithoutextensivediscussion,whichisreservedforthefollowingsection.Thisisappropriatewhenthedataintheearlypartsdonotneedtobeinterpretedextensivelytounderstandwhythelaterexperimentsweredone.Inotherpapers,resultsaregiven,andthentheyareinterpreted,perhapstakentogetherwithotherfindingsnotinthepaper,soastogivethelogicalbasisforlaterexperiments.

Thefourthsectionisthe Discussion.Thissectionservesseveralpurposes.First,thedatainthepaperareinterpreted;thatis,theyareanalyzedtoshowwhattheauthorsbelievethedatashow.Anylimitationstotheinterpretationsshouldbeacknowledged,andfactshouldclearlybeseparatedfromspeculation.Second,thefindingsofthepaperarerelatedtootherfindingsinthefield.Thisservestoshowhowthefindingscontributetoknowledge,orcorrecttheerrorsofpreviouswork.Asstated,someoftheselogicalargumentsareoftenfoundintheResultswhenitisnecessarytoclarifywhylaterexperimentswerecarriedout.AlthoughyoumightarguethatinthiscasethediscussionmaterialshouldbepresentedintheIntroduction,moreoftenyoucannotgraspitssignificanceuntilthefirstpartofResultsisgiven.

Finally,papersusuallyhaveashort Acknowledgements section,inwhichvariouscontributionsofotherworkersarerecognized,followedbya Reference listgivingreferencestopapersandotherworkscitedinthetext.

Papersalsocontainseveral Figures and Tables.Thesecontaindatadescribedinthepaper.Thefiguresandtablesalsohavelegends,whosepurposeistogivedetailsoftheparticularexperimentorexperimentsshownthere.Typically,ifaprocedureisusedonlyonceinapaper,thesedetailsaredescribedinMaterialsandMethods,andtheFigureorTablelegendrefersbacktothatdescription.Ifaprocedureisusedrepeatedly,however,ageneraldescriptionisgiveninMaterialsandMethods,andthedetailsforaparticularexperimentaregivenintheTableorFigurelegend.

Variationsontheorganizationofapaper

Inmostscientificjournals,theaboveformatisfollowed.Occasionally,theResultsandDiscussionarecombined,incasesinwhichthedataneedextensivediscussiontoallowthereadertofollowthetrainoflogicdevelopedinthecourseoftheresearch.Asstated,insomejournals,MaterialsandMethodsfollowstheDiscussion.Incertainolderpapers,theSummarywasgivenattheendofthepaper.

Theformatsfortwowidely-readjournals, Science and Nature,differmarkedlyfromtheaboveoutline.Thesejournalsreachawideaudience,andmanyauthorswishtopublishinthem;accordingly,thespacelimitationsonthepapersaresevere,andtheproseisusuallyhighlycompressed.Inbothjournals,therearenodiscretesections,exceptforashortabstractandareferencelist.InScience,theabstractisself-contained;in Nature,theabstractalsoservesasabriefintroductiontothepaper.Experimentaldetailsareusuallygiveneitherinendnotes（for Science）orFigureandTablelegendsandashortMethodssection（in Nature）.Authorsoftentrytocircumventlengthlimitationsbyputtingasmuchmaterialaspossibleintheseplaces.Inaddition,anincreasinglycommonpracticeistoputasubstantialfractionoftheless-importantmaterial,andmuchofthemethodology,intoSupplementalDatathatcanbeaccessedonline.

Manyotherjournalsalsohavelengthlimitations,whichsimilarlyleadtoaneedforconciseness.Forexample,the ProceedingsoftheNationalAcademyofSciences （PNAS）hasasix-pagelimit;Cell severelyeditsmanypaperstoshortenthem,andhasashortwordlimitintheabstract;andsoon.

Inresponsetothepressuretoeditandmakethepaperconcise,manyauthorschoosetocondenseor,moretypically,omitthelogicalconnectionsthatwouldmaketheflowofthepapereasy.Inaddition,muchofthebackgroundthatwouldmakethepaperaccessibletoawideraudienceiscondensedoromitted,sothattheless-informedreaderhastoconsultareviewarticleorpreviouspaperstomakesenseofwhattheissuesareandwhytheyareimportant.Finally,again,authorsoftencircumventpagelimitationsbyputtingcrucialdetailsintotheFigureandTablelegends,especiallywhen（asin 

PNAS）thesearesetinsmallertype.Fortunately,therecentwidespreadpracticeofputtingless-criticalmaterialintoonlinesupplementalmaterialhaslessenedthepressuretocompresscontentsodrastically,butitisstillaproblemforolderpapers.

Backtooutline

2.Readingascientificpaper

Althoughitistemptingtoreadthepaperstraightthroughasyouwoulddowithmosttext,itismoreefficienttoorganizethewayyouread.Generally,youfirstreadtheAbstractinordertounderstandthemajorpointsofthework.Theextentofbackgroundassumedbydifferentauthors,andallowedbythejournal,alsovariesasjustdiscussed.

OneextremelyusefulhabitinreadingapaperistoreadtheTitleandtheAbstractand,beforegoingon,reviewinyourmindwhatyouknowaboutthetopic.Thisservesseveralpurposes.First,itclarifieswhetheryouinfactknowenoughbackgroundtoappreciatethepaper.Ifnot,youmightchoosetoreadthebackgroundinareviewortextbook,asappropriate.

Second,itrefreshesyourmemoryaboutthetopic.Third,andperhapsmostimportantly,ithelpsyouasthereaderintegratethenewinformationintoyourpreviousknowledgeaboutthetopic.Thatis,itisusedasapartoftheself-educationprocessthatanyprofessionalmustcontinuethroughouthis/hercareer.

Ifyouareveryfamiliarwiththefield,theIntroductioncanbeskimmedorevenskipped.Asstatedabove,thelogicalflowofmostpapersgoesstraightfromtheIntroductiontoResults;accordingly,thepapershouldbereadinthatwayaswell,skippingMaterialsandMethodsandreferringbacktothissectionasneededtoclarifywhatwasactuallydone.AreaderfamiliarwiththefieldwhoisinterestedinaparticularpointgivenintheAbstractoftenskipsdirectlytotherelevantsectionoftheResults,andfromtheretotheDiscussionforinterpretationofthefindings.Thisisonlyeasytodoifthepaperisorganizedproperly.

Codewords

Manypaperscontainshorthandphrasesthatwemightterm'codewords',sincetheyhaveconnotationsthataregenerallynotexplicit.Inmanypapers,notalltheexperimentaldataareshown,butreferredtoby"（datanotshown）".Thisisoftenforreasonsofspace;thepracticeisacceptedwhentheauthorshavedocumentedtheircompetencetodotheexperimentsproperly（usuallyinpreviouspapers）.Twoothercodewordsare"unpublisheddata"and"preliminarydata".Theformercaneith