a-survey-study-(调查研究)

Chapter6AsurveystudyLiuFei201312004目录abriefdescriptionquestionnairescalingtechniquesadministeringthequestionnaireselectingsubjects6.1abriefdescription•Asurveystudyusuallyinvolvesarelativelylargesampleandisdescriptiveinnature.•Iteitherdescribes①thefeaturesofafewvariablesor②therelationbetweentwoormorevariables.•调查研究指的是采用观察、问卷、访谈等方式从一个样本那里收集资料，通过对资料的分析来认识社会现象及其规律。（孙三军周晓岩，2011，p.22）6.1abriefdescription•Cross-sectional:dataarecollectedfromdifferentpeopleatonepointintime.•Longitudinal:dataarecollectedfromthesamegroupofpeopleatdifferentpointsoveraperiodoftime.6.1abriefdescription•Tocollectdatainasurveystudy,weareexpectedtoundertakeseveraltasks:⑴constructinganinstrumentfordata-collection;⑵selectingsubjectsforthestudy;⑶collectingdataeitherpersonallyorthroughcorrespondence.6.2instrumentdesigning:questionnaire•Thequestionnaireisoneofthemostwidelyusedtechniquesforcollectingquantitativedata.Advantages:lesstime＆lessexpense.Disadvantages:lackingoftheoryinthesensethattherearenoscientificprinciplesthatguaranteeanoptimaloridealquestionnaire.6.2.1criteriaforagoodquestionnaire•Anygoodquestionnaireshouldatleastmeetthefollowingtwocriteria:HavinghighinternalvalidityTakingaprofessionaloutlookCriteria1:Havinghighinternalvalidity•Itemsinthequestionnairemustmeasurethevariableswereallywanttoinvestigate.•效度即有效性，系指测量工具或手段能够准确测出所需测量的事物的程度。（孙三军周晓岩，2011，p.131）•内部效度表示实验所提供的自变量和因变量之间因果关系明确程度的一种指标，通常讲就是研究的有效性。•外部效度是指研究结果能够一般化和普遍适用到样本来自的总体和到其它的总体中的程度，即研究结果和变量条件、时间和背景的代表性和普遍适用性。※Aboutvalidity•效度即有效性，系指测量工具或手段能够准确测出所需测量的事物的程度。•效度是科学的测量工具所必须具备的最重要的条件。在社会测量中，对作为测量工具的问卷或量表的效度要求较高。比如：•用一份英文问卷调查国内大一学生对某事物的看法，由于被试英文水平不一，所以问卷所测的与其说是被试对某一事物的看法，不如说是他们的英文水平，该问卷也就缺乏效度。※Aboutreliability•信度即可靠性，系指采用同样的方法对同一对象重复测量时所得结果的一致性程度，多以相关系数表示。•信度分析方法多采用cronbach’sα信度系数（用SPSS软件计算）。cronbach’sα信度系数法测的是量表的内在一致性，需要对一组被试测试一次，所得系数大小要看两个因素：量表中题目的数量题目分值差异与总分值差异的比率•量表中题目越多，题目分值差异相对于总分值差异越小，所得系数就越大。※Aboutreliability•一般情况下，相关系数为0.70以上属于高相关，不过多数研究者希望相关系数在0.80以上。相关系数在0.70以下则表明依据该量表所得分数并不可靠。•要提高相关系数，方法有：剔除相关系数低的问题增加问题数目•但是问题太多会增加被试的负担，所以研究者在决定量表长度时需要权衡利弊Criteria2:Takingaprofessionaloutlook•Agoodquestionnairewithaprofessionaloutlookshouldnotcontainthefollowingproblems:Youshouldavoidclippingorstaplingpaperstogetherifaquestionnairecontainsmorethan5pages.Onequestionshouldnotbeprintedacrosstwopagessincesplitquestionnairemayleadrespondentstothinkthatthequestionhasendedattheendofapage.Weshouldavoidcrowdingquestionstogethertomakequestionnairestolookshorter.6.2.2Questionnairedesignprocess•Thedesignofaquestionnairecanbedescribedintermsofaseriesofstepsthatinclude:①Selectingthemodeforcollectingquestionnairedata;②Specifyingwhatkindofdatayouintendtocollect;③Determiningwhetherquestionnaireitemswillbeprocessedindividuallyorascategories;④Decidingthecontentofindividualitems;⑤Choosingthequestionstructure;⑥Selectingthequestionwording;⑦Arrangingthequestionsinproperorder;⑧Decidingtheformatofthequestionnaire;⑨Pretestingthequestionnaires;Step1:Selectingthemodesofadministration•Themodeofadministrationmeansinwhichwayquestionnairesaredistributedtotherespondents.BypersonBymailYourchoiceofamodewilldependonyourownconsiderationofcost,convenienceandthenatureofthequestionsyouareasking.Step2:Specifyingwhatkindofdatayouintendtocollect•Therearebasicallyfivetypesofinformation:①Experience/behavior②Opinion③Feeling④Knowledge/abilities⑤backgroundStep2:SpecifyingwhatkindofdatayouintendtocollectAttention•Veryoften,noviceresearcherscannotclearlydifferentiatebeliefitemsfrombehavioritems.Beliefitems:“IthinklearninganL2wellrequirespainstakingeffort.”Behavioritems:“IputalotofeffortintolearninganL2.”Step3:Determiningthewayyouprocessthequestionnairedata•DatabeanalyzedintermsofIndividualitemsorintermsofcategories•Ifthebasicunitinthedataanalysisisanindividualitem,suchasdescribingthefrequencyandthemeanofoneisolateditem,thensuchaquestionnaireiscalledanindividual-item-basedquestionnaire.•Ifthedata-analysisistobebasedoncategories,eachofwhichcontainsafewquestionnaireitems,youmustestablishconceptualcategorieseitherbyatop-downapproachorbyabottom-upapproach.Step3:Determiningthewayyouprocessthequestionnairedata•Byatop-downapproachyouconstructconceptualcategoriesbasedonlogicalargumentsorexistingtheoriesbeforeyoudesignspecificitems.•Byabottom-upapproach,yousimplywritedownwhateveritemsoccurinyourmind.Step4:DecidingonthecontentofindividualitemsStep1:•Ifaquestionnaireisintendedtocontainseveralcategories,youhavetoconsider①whethertheitemsdesignedcoverallthecategoriesand②whethereachcategorycontainsanadequatenumberofitems.•Inordertoanswerthesetwoquestions,youhavetomakeupatablewhichhastwocolumns.Inonecolumn,youwritedownthecategoriesandintheothercolumn,youputdownthesequentialnumberofeachquestionnaireitem.(p104table6.2)Step4:DecidingonthecontentofindividualitemsCategorynameItemNo.Form-focusedbeliefs2,4,5,7,9Meaning-focusedbeliefs1,3,6,8,10Form-focusedstrategies11,13,16,18,20Meaning-focusedstrategies12,14,15,17,19Step2:•Youhavetoreadtheitemswiththesamecategoryagainandagaintomakesurethatalltheitemslogicallygotogether.Step3:•Finallyyouhavetoch