Willem Mertens · Amedeo Pugliese Jan Recker Quantitative Data Analysis A Companion for Accounting and Information Systems Research Quantitative Data Analysis Willem Mertens (cid:129) Amedeo Pugliese (cid:129) Jan Recker Quantitative Data Analysis A Companion for Accounting and Information Systems Research WillemMertens AmedeoPugliese QUTBusinessSchool Dept.ofEconomicsandManagement QueenslandUniversityofTechnology UniversityofPadova Brisbane,Queensland Padova,Italy Australia SchoolofAccountancy QueenslandUniversityofTechnology Brisbane,Queensland Australia JanRecker QUTBusinessSchool QueenslandUniversityofTechnology Brisbane,Queensland Australia ISBN978-3-319-42699-0 ISBN978-3-319-42700-3 (eBook) DOI10.1007/978-3-319-42700-3 LibraryofCongressControlNumber:2016954196 #SpringerInternationalPublishingSwitzerland2017 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpartof the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilarmethodologynowknownorhereafterdeveloped. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexempt fromtherelevantprotectivelawsandregulationsandthereforefreeforgeneraluse. Thepublisher,theauthorsandtheeditorsaresafetoassumethattheadviceandinformationinthis book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, express or implied, with respect to the material contained hereinorforanyerrorsoromissionsthatmayhavebeenmade. Printedonacid-freepaper ThisSpringerimprintispublishedbySpringerNature TheregisteredcompanyisSpringerInternationalPublishingAGSwitzerland Preface Quantitative Data Analysis for Accounting and Information Systems Research guides postgraduate research students and early career researchers in choosing andexecutingappropriatedataanalysismethodstoanswertheirresearchquestions. Itsupportsresearcherswhentheyareplanningtocollectdataandwhentheyhave datathattheywanttoanalyze.Usingavarietyofexamplesandlaylanguage,this bookprovideshands-onguidelineson(1)whentousewhichdataanalysismethod, (2) what kind of data and data structure are required for each of the methods, (3) what each method does and how the methods should be used, and (4) how to reportresults. This book is not intended to provide an exhaustive overview of data-analysis methods,nordoesitexplainthemethodsindepth.Instead,itguidesresearchersin applyingtherightmethodsintherightway.Moreskilledresearcherscanalsouse the book to refresh their knowledge or as a checklist to avoid skipping important steps. It explains the most commonly used methods in an intuitive and hands-on way, pointing out more advanced resources along the way. As such, it does not aspiretocompetewithmanualslikethoseofStevens[1],Field[2],orCrawley[3]. Wearenotstatisticiansbutresearcherswhoapplystatistics,1sothebookcoversthe issues that commonly affect others like us, who are engaging in quantitative empiricalresearch. QuantitativeDataAnalysisforAccountingandInformationSystemsResearchis thebookwewouldhavelikedtohavehadasasupportinourownresearch.Every chapter provides an unintimidating starting point for building your data-analysis skills,theinformationrequiredtorunthemostcommonanalysesandreportthem, andpointerstomoreextensiveresources.Attheriskofsayingthingsthatmaynot be entirely true in the purest statistical sense, we try to keep the language of this bookassimpleaspossible.Assuch,thebookisbriefandwritteninalanguagethat we hope everyone can understand—from students to researchers to people who wishtostudytheorganizationsinwhichtheywork.Ourgoalistohelpyouconduct academicresearchofhighqualityanddotherightthingsright—nottomakeyoua 1If youarea statistician orsimplymore observantthanwe are, weinvite you totell usifyou identifyanerror. v vi Preface statisticsexpert—sothisbookisnotaboutstatisticsbutaboutapplyingstatisticsto theresearchquestionsthatkeepyouawakeatnight.(Wedoubtthesequestionsare aboutcollinearity,butiftheyare,thismaynotbethebookyouarelookingfor.) In brief, this book is a software-independent starting point for answering the question:WhatmethodsdoIusetoanswermyresearchquestionsandhow? Wehopeyouhavefun! Brisbane,QLD,Australia WillemMertens Padova,Italy AmedeoPugliese JanRecker References 1. StevensJP(2009)Appliedmultivariatestatisticsforthesocialsciences.TaylorandFrancis, LLC,London 2. FieldAP(2013)DiscoveringstatisticsusingIBMSPSSstatistics,andsexanddrugsandrock ‘n’roll,4thedn.Sage,London 3. CrawleyMJ(2013)TheRbook,2ndedn.Wiley,WestSussex Acknowledgments Manypeoplehavecontributedtothedevelopmentofthisbookanditscontent.We are grateful to everyone who helped us discover, explain, and write down our understandingof what matters inusing statisticsfor dataanalysis. Although there aremanyofyououtthere,wewouldliketothankafewinparticular. First, we are grateful for the support of the QUT’s School of Management, School of Accountancy and Information Systems School, for supporting us in the development and conduct of the Advanced Data Analysis workshop series; this book would not have been possible without it. Special thanks go to Professor Michael Rosemann and Professor Peter Green for their inspiring entrepreneurial spirit, flexibility, and support. Second, we are grateful that so many of our colleagues and students attended these workshops and discussed and challenged ourunderstandingofdataanalysismethodsandthewaywetaughtthem. Finally, the ones that contributed—or perhaps suffered—most are our lovely wives,Laura,Claudia,andLaura.Thankyouforyoursupport,yourpatience,and for sharing some of our headaches. You make our lives 89% more enjoyable (p<.001,[75–100]). September2016 WillemMertens AmedeoPugliese JanRecker vii Contents 1 Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1 1.1 IntroductiontotheBasics. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 1.2 NavigatingtheWorldofStatistics—AndThisBook. . . . . . . . . . 3 1.3 WhatThisBookDoesNotCover. . . . . . . . . . . . . . . . . . . . . . . 5 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6 2 ComparingDifferencesAcrossGroups. . . . . . . . . . . . . . . . . . . . . . 7 2.1 OneorTwoGroups. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7 2.2 MorethanTwoGroups:One-WayANOVA.. . . . .. . . .. . . . .. 10 2.3 MorethanTwoGroupingVariables:FactorialANOVA. . . . . . . 12 2.4 MorethanOneDependentVariable:MultivariateANOVA. . . . 14 2.5 MoreAdvancedModels:CovarianceandRepeatedMeasures. . . 16 2.6 WhentoUseGroupComparisons. . . . . . . . . . . . . . . . . . . . . . . 17 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 3 Assessing(Innocuous)Relationships. . . . . . . . . . . . . . . . . . . . . . . . 21 3.1 WhatAreRegressionModels?. . . . . . . . . . . . . . . . . . . . . . . . . 21 3.2 WhenDoWeUseRegressionModels?. . . . . . . . . . . . . . . . . . . 24 3.3 HowDoWeExamineRegressionModels?. . . . . . . . . . . . . . . . 26 3.4 HowDoWeReportRegressionAnalyses?. . . . . . . . . . . . . . . . . 31 3.5 WhatIf.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 36 4 ModelswithLatentConceptsandMultipleRelationships: StructuralEquationModeling. . . . . . . . . . . . . . . . . . . . . . . . . . . . 37 4.1 WhatAreStructuralEquationModels?. . . . . . . . . . . . . . . . . . . 37 4.2 WhenDoWeUseStructuralEquationModels?. . . . . . . . . . . . . 41 4.3 HowDoWeExamineStructuralEquationModels?. . . . . . . . . . 43 4.4 HowDoWeReportStructuralEquationModelAnalyses?. . . . . 49 4.5 WhatIf.... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57 5 NestedDataandMultilevelModels:HierarchicalLinearModeling... 61 5.1 WhatAreHierarchicalLinearModels?. . . . . . . . . . . . . . . . . . . 61 5.2 WhenDoWeUseHLMs?. . . . . . . . . . . . . . . . . . . . . . . . . . . . 63 ix x Contents 5.3 HowDoWeInvestigateHLMs?. . . . . . . . . . . . . . . . . . . . . . . . 64 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 6 AnalyzingLongitudinalandPanelData. . . . . . . . . . . . . . . . . . . . . 73 6.1 WhatAreLongitudinalandPanelData?. . . . . . . . . . . . . . . . . . 73 6.2 ClusteringasaWaytoDealwithNestedness. . . . . . . . . . . . . . . 77 6.3 WhichModelsCanWeUsetoAnalyzeLongitudinalData?. . . . 81 6.4 EstimatingandReportingFixed-EffectsandRandom-Effects Models.. . . .. . . .. . . .. . . .. . . . .. . . .. . . .. . . .. . . .. . . .. 84 6.5 WhentoUseOLS,Fixed-Effects,andRandom-EffectsModels. . . 92 6.6 FinalRemarks,SuggestionsandYourBestTool:Thinking!. . . . 96 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 7 Causality:EndogeneityBiasesandPossibleRemedies. . . . . . . . . . 99 7.1 YourResearchQuestionIsCausal:WhatDoesthatMean?. . . . . 100 7.2 Self-SelectionandEndogeneity. . . . . . . . . . . . . . . . . . . . . . . . . 105 7.3 SpecifyingOLSModelstoMinimizeEndogeneityConcerns. . . 107 7.4 MoreComplexDesignstoSupportCausalClaims. . . . . . . . . . . 111 7.5 SomeCaveatsandLimitations. . . . . . . . . . . . . . . . . . . . . . . . . 132 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 133 8 HowtoStartAnalyzing,TestAssumptionsandDeal withthatPeskyp-Value. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 135 8.1 Structuring,Cleaning,andSummarizingData. . . . . . . . . . . . . . 136 8.1.1 StructuringData. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136 8.1.2 CleaningData. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137 8.1.3 ExploringData:SummaryStatisticsandVisualization. . . 142 8.2 TestingAssumptions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 8.2.1 IndependenceofObservations. . . . . . . . . . . . . . . . . . . . 143 8.2.2 Normality. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 147 8.2.3 HomogeneityofVarianceandHomoscedasticity. . . . . . . 148 8.2.4 Linearity. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 149 8.2.5 WhatifAssumptionsAreViolated?. . . . . . . . . . . . . . . . 150 8.3 MindfullyInterpretingStatistics:TheCaseofthep-Value. . . . . 152 References. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155 9 KeepingTrackandStayingSane. . . . . . . . . . . . . . . . . . . . . . . . . . 157 Index. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 161 1 Introduction Data, data, data. More data is available to us now than ever before. As work and privateactivitiesareincreasinglyfacilitatedandenactedbyourdigitaldevices,we leave traces that can be picked up and analyzed anytime and anywhere. Data collected through surveys, archives, and experiments also remain relevant, as digital traces do not necessarily reflect perceptions, attitudes, and intentions. What also has not changed is that data is meaningless until it is analyzed. That is what this book is about: analyzing data. More precisely: analyzing quantitative data.Numbers. Dataanalysisisaniterativeprocessofmanipulatingandinterpretingnumbersto extractmeaningfromthem—answerresearchquestions,testhypotheses,orexplore meaningsthatcanbederivedinductivelyfromthedata.Thisexplorationisthefirst stepofanydataanalysis:werunafewbasicmanipulationsandteststosummarize the data in meaningful statistics, such as means and standard deviations; we visualize the data; and we try to improve our understanding of the information in thedata. Of course, before you can start analyzing, you need to obtain data and have a rough idea of the meaning you want to extract through analysis. Therefore, every chapterbrieflydiscusseswhenandwhyyoumaywanttousethemethodsdiscussed inthatchapter,including thetypeofquestionstypicallyansweredandthetypeof data analyzed using that method. We hope that this approach will help you understand how theory, research designs, research questions, data, and analysis dependononeanother.Thecredibilityofdataisderivedfromtheresearchdesign; andthecredibilityofdataanalysisisderivedfromitsgroundingintheory. Beforewestartdiscussingdataanalysismethods,wewanttosummarizethekey conceptsusedinthisbookandgiveyouaroadmapforusingthebookandchoosing therightanalysismethod. #SpringerInternationalPublishingSwitzerland2017 1 W.Mertensetal.,QuantitativeDataAnalysis,DOI10.1007/978-3-319-42700-3_1
Description: