haihongyuan.com
海量文库 文档专家
全站搜索:
您现在的位置:首页 > 小学教育 > 小学作文小学作文

如何论文写作 不错的文章 老师推荐

发布时间:2013-12-14 16:46:42  

COMMENTARY

SpinandBoastinginResearchArticles

S

OMEAUTHORSEXAG-

geratetheimportanceoftheirresearchandunfairlydenigrateotherstudies.Thisoccurs

onlyinaminorityofarticleswere-viewbutisfrequentenoughthatwehavecollectedexamplesandgroupedthemintocategories.Weconfesstocommittingsomeoftheseliterarysinsourselves;improvingwritingisalife-longprocess.

EXAGGERATINGTHEIMPORTANCEOFTHETOPICINTHEINTRODUCTION

Examplesinclude:“[X]hasreachedanalarmingproportion,”“anawe-inspiringnumber,”and“drasticin-crease.”Thesefeverishwordslackscientificmeaning.Instead,giverel-evantquantitiesforaparticularplaceandtime:countsofevents,inci-dencerates,orprevalenceesti-mates.Considerdescribinghowthesequantitieshavechangedovertime.Letthereaderdecidewhetheranythingisalarming,awesome,ordrastic.Writingthatsomethingis“acriticalpriority,”eveniftrueinsomesense,isjustaneditorialopinionun-likelytopersuade.Referringto“theobesityepidemic”isacliche′.Hack-neyedphrasesdonotmakethewriterappearthoughtful,arebor-ingforthereader,andtakeupspace.Considerwhetherthereaderneedstoonceagainhearthatobesityiscommon,diabetesisincreasing,andthatthecostofmedicalcareisaproblem.Wethinknot.Research-ersshoulduseaminimumofadjec-tivestodescribetheirtopic.UNFAIRLYDISPARAGINGPREVIOUSRESEARCHTopromotetheirwork,authorshavedeclareda“crisisofcredibility”anda“largeresearchgap.”Somedis-misspaststudiesforthefollowingspeciousreasons:

v“Inconsistent”or“mixed”re-sults:Thesefuzzywordsexplainnothing.Wedostudiesinfinitesamples,sosomedifferenceinre-sultsisalmostinevitable.Ifastatis-ticaltestofhomogeneityindicatesthedifferenceismorethanex-pected,anewstudyisunlikelytochangethat.Considermeta-analyticmethodstodescribethevariationandseekexplanations,suchasdissimilarstudypopulationsorde-signs.

v“Methodologicflaws”:Thesewordsarenothelpfulbecausees-sentiallyallstudieshaveflaws.Thenewstudywillhaveitsownflaws.Ifthereisevidencethatapriorstudyisbiased,byallmeansshowthis.Alessthanoptimaldesignisnotproofofbias.

v“Smallsamplesize”:Somesmallstudiesprovideusefulevidence.Thefirstreportthatdiethylstilbestroltakenbypregnantwomenwasre-latedtovaginalcancerintheiroff-springhadonly8casesand32con-trols.1Ifallelseisequal,alargerstudywillproduceanarrowercon-fidenceintervalaroundanestimatecomparedwithasmallerstudy.Butallelsemaynotbeequal;estimatesfromthelargerstudymightbemorebiased.Ifseveralsmallstudieshavebeendone,poolingtheirestimatesusingmeta-analysismayresultinaconfidenceintervalasnarrowasthatfromalargerstudy.

v“Retrospective”:Inaccuratere-callisapotentialsourceofbias,notevidenceofbias.Thestudiesthatlinkedpronesleeppositiontosud-deninfantdeathsyndromewereretrospective.

Oneapproach(amongothers)istosaynothingorlittleaboutpastworkintheintroduction.Forex-ample,“Theassociationofexpo-sureXwithoutcomeYhasbeenes-timatedin5case-controlstudies[citationshere]whichusedsubjectrecalltomeasureexposureafterout-

comeswereknown.Wealsousedacase-controldesign,butobtainedex-posuredatafromrecordscreatedpriortotheoutcome.”Inthedis-cussionsummarizeeachstudyandcompareitsestimatetoyours:“Astudyof125Norwegianchildrenre-portedthatthoseexposedtoX,com-paredwiththosenotexposed,hadYmoreoften:adjustedriskratio,2.00(95%CI,0.96-4.16).Thises-timate,basedonrecallofexposure,waslittledifferentfromourad-justedriskratioof1.9(95%CI,1.31-2.75)basedonpastrecordsofex-posure.”Thiswordingsuggeststhatrecallisapotentialsourceofbias,butthereislittleevidencethatitac-tuallyproducedbias.

USINGWORDSTOCONVINCE,

NOTILLUMINATESomeauthorsclaimclarity:“Ourre-sultsclearlyshow,”“theseanalysesprovideclearanswers,”“[X]wasclearlyaffected,”“webelieveourstudymakesitclear.”Otherssug-geststrength:“Our...analy-sis...providesstrongevidence,”“thisstudyshowsthat[X]isveryef-fective,”“therobustassociation.”Re-searchersshouldpresentevidencesoothersmayunderstandandjudgeit.Iffindingsarelucidandpowerful,wordstopersuadearesuperfluous.Ifotherwise,thesewordsjustin-vitethereadertoconsiderwhynei-theristrue.

BOASTING

Boastsofbeingfirstarecommon.Someareinadvertentlyamusingbe-causetheyhavesomanyqualifiers,likebraggingaboutbeingtheold-estleft-handedpersontowalkback-warduptheWashingtonMonu-ment.Ifthestudyisfirst,lackofcomparisontopreviousworkinthecommentsectionwillmakethisap-parent.Ifthecrowingaboutprior-ityiswrong,notrareinourexperi-

ARCHPEDIATRADOLESCMED/VOL166(NO.12),DEC

2012

1099

WWW.ARCHPEDIATRICS.COM

?2012AmericanMedicalAssociation.Allrightsreserved.

ence,thosewhoweretrulyfirstwillnotbeyourfriendsandyourevealignoranceaboutpriorwork.

Otherboastsinclude:“thefind-ings...openupanewfrontier,”“thisstudyrepresentsthefirstun-equivocaldemonstration,”“largestsamplesize,”“strongnewevi-dence,”“oneofthemostde-tailed...studies,”“thisstudypro-videsnewinsights,”“ourcomprehensiveassessment.”Resistimmodestchest-thumping.ANEXAMPLEOFEXCELLENT

PRESENTATIONWecanlearnfromgoodexamples.Resultsofarandomizedcontrolledtrialofscreeningforovariancan-cerwerepublishedin2011.2Thein-troductionhasoneparagraphwithrelevantfactsaboutovariancancerandanotheraboutscreeningtrials,withouteditorialhype.Thisstudyrandomized34253womentospe-cialscreeningand34304tousualcare.Theauthorsneverbragthatthestudywaslargeorfirst.Thecom-

mentsectionprovideshelpfulinter-pretation,equitabledescriptionof2otherscreeningtrials,anddis-cusseslimitations.Wedetectedonlyonedispensableadjective;invasiveprocedurestriggeredbyfalse-positivetestresultswerecalled“ase-riousconcern.”Weagree,but“se-rious”wasunneeded.Theconclusioneschewedbothpolicyad-vice3,4orboringsuggestionsformoreresearch.5Thismanuscriptavoidedbothspinandboasting.

Writingforscientificjournalsshouldbeascleanandconciseaspossible.Leavespinandboastingtoothers.

PeterCummings,MD,MPHFrederickP.Rivara,MD,MPHPublishedOnline:October22,2012.doi:10.1001/archpediatrics.2012.1461

AuthorAffiliations:DepartmentofEpidemiologyandHarborviewIn-juryPreventionandResearchCen-ter(DrsCummingsandRivara);andDepartmentofPediatrics(DrRi-vara),UniversityofWashington,and

SeattleChildren’sHospital,Seattle,Washington.

Correspondence:DrCummings,DepartmentofEpidemiology,Uni-versityofWashington,250Grand-viewDr,Bishop,CA93514(peterc@uw.edu).

FinancialDisclosure:Nonere-ported.

REFERENCES

1.HerbstAL,UlfelderH,PoskanzerDC.Adenocarci-nomaofthevagina:associationofmaternalstil-bestroltherapywithtumorappearanceinyoungwomen.NEnglJMed.1971;284(15):878-881.2.BuysSS,PartridgeE,BlackA,etal;PLCOProjectTeam.Effectofscreeningonovariancancermor-tality:theProstate,Lung,ColorectalandOvarian(PLCO)CancerScreeningRandomizedControlledTrial.JAMA.2011;305(22):2295-2303.

3.WeissNS.Policyemanatingfromepidemiologicdata:whatistheproperforum?Epidemiology.2001;12(4):373-374.

4.CummingsP.Policyrecommendationsinthedis-cussionsectionofaresearcharticle.InjPrev.2007;13:4-5.doi:10.1136/ip.2007.015156.

5.RivaraFP,CummingsP.WritingforpublicationinArchivesofPediatrics&AdolescentMedicine[editorial].ArchPediatrAdolescMed.2001;155(10):1090-1092.

ARCHPEDIATRADOLESCMED/VOL166(NO.12),DEC

2012

1100

WWW.ARCHPEDIATRICS.COM

?2012AmericanMedicalAssociation.Allrightsreserved.

网站首页网站地图 站长统计
All rights reserved Powered by 海文库
copyright ©right 2010-2011。
文档资料库内容来自网络,如有侵犯请联系客服。zhit326@126.com