Jared A. Linebach · Brian P. Tesch Lea M. Kovacsiss Nonparametric Statistics for Applied Research Nonparametric Statistics for Applied Research Jared A. Linebach (cid:129) Brian P. Tesch Lea M. Kovacsiss Nonparametric Statistics for Applied Research JaredA.Linebach BrianP.Tesch ClearwaterChristianCollege SuffolkUniversity Clearwater,FL,USA Dover,NewHampshire,USA LeaM.Kovacsiss EastCanton,OH,USA ISBN978-1-4614-9040-1 ISBN978-1-4614-9041-8(eBook) DOI10.1007/978-1-4614-9041-8 SpringerNewYorkHeidelbergDordrechtLondon LibraryofCongressControlNumber:2013950181 ©SpringerScience+BusinessMediaNewYork2014 Thisworkissubjecttocopyright.AllrightsarereservedbythePublisher,whetherthewholeorpart of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation,broadcasting,reproductiononmicrofilmsorinanyotherphysicalway,andtransmissionor informationstorageandretrieval,electronicadaptation,computersoftware,orbysimilarordissimilar methodologynowknownorhereafterdeveloped.Exemptedfromthislegalreservationarebriefexcerpts inconnectionwithreviewsorscholarlyanalysisormaterialsuppliedspecificallyforthepurposeofbeing enteredandexecutedonacomputersystem,forexclusiveusebythepurchaserofthework.Duplication ofthispublicationorpartsthereofispermittedonlyundertheprovisionsoftheCopyrightLawofthe Publisher’s location, in its current version, and permission for use must always be obtained from Springer.PermissionsforusemaybeobtainedthroughRightsLinkattheCopyrightClearanceCenter. ViolationsareliabletoprosecutionundertherespectiveCopyrightLaw. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publicationdoesnotimply,evenintheabsenceofaspecificstatement,thatsuchnamesareexempt fromtherelevantprotectivelawsandregulationsandthereforefreeforgeneraluse. While the advice and information in this book are believed to be true and accurate at the date of publication,neithertheauthorsnortheeditorsnorthepublishercanacceptanylegalresponsibilityfor anyerrorsoromissionsthatmaybemade.Thepublishermakesnowarranty,expressorimplied,with respecttothematerialcontainedherein. Printedonacid-freepaper SpringerispartofSpringerScience+BusinessMedia(www.springer.com) WearemostgratefultoDr.DebraBekerian,Ph.D., for her unwavering commitment to us and the process. Without her guidance and encouragement, this would never have been possible. To you, we dedicate this work. Preface Ihavebeenworkingasanappliedpsychologistformanyyears,andthereareafew things that have consistently stood out, for me at least, in the course of my experiences. Possibly the single, most constant “truth” is that human behavior is messy. It’s messy in all sorts of interesting ways, and most of the time, people’s messiness also messes with any type of inference you can make about their behavior.So,peoplemaynotbehave,asagroup,inanormallydistributedfashion, orasa“singlehumpedcamel,”astheauthorssayinthisbook. In fact, applied research is messy. For example, take how you get participants. You put out feelers, such as links on various websites; you advertise you need participantsforastudyonwhateveritisyouhappentobestudying.Theindividual decidestorespondornot—astheresearcher,youprettymuchhavetotakewhoyou canget.Youalsodon’talwayshavetheopportunitytousemeasurementsthatyou’d like.So,youmaybereducedtoaskingyes/noquestions,simplybecauseyoucannot passanethicsboard,peoplewouldn’tanswerthequestionsyoureallywanttoask orboth. And, ofcourse, whenyou’re dealing withmessybehavior,there isn’t always a nice, tidy way of determining whether you’ve found anything significant. That’s right;I’mtalkingaboutparametricstatistics.Intherealworld,theparametersareso oftenviolatedthatyouneedtofindanotherway. To this end, nonparametric statistics offer a delightful smorgasbord of alterna- tivesfromwhichtosample.Nomatterhowsloppy,nomatterhowimprecise,and nomatterhowadhocthebehavioralmeasurement,nonparametricstatisticsprom- ise some light at the end of the tunnel, a way to assess whether your findings are potentiallypointingtosomethingsignificant. Whilethereareanumberoftextbooksonnonparametricstatistics,noneofthem offerswhatthisbookdoes.Thisbookisuniqueinanumberofways.Forone,the text provides a context for statistical questions: there are applied problems that drivetheanalyses,andtheproblemsarelinkedtoeachothersothatthereadergetsa real appreciation of how applied science works. The data set used by the book is consistent, too. What this means is that the reader is allowed to become familiar, and confident, with one set of numbers, rather than changing each data set with a vii viii Preface new statistical test (the traditional statistics book approach). Also unusual and highly valuable is the decision tree for tests of differences and of association. I am convinced that these trees will facilitate the problem solving process for studentsofpsychologyaswellasseasonedresearchers. The book also departs from the standard in that it provides the reader with a narrative of real people, doing real things and interacting with each other in real ways. The issues are real, the consequencesserious. The reader is introduced toa contextinwhichstatisticsgetapplied,andasaconsequence,therationaleforusing a test is grounded in an understandable example. This is in stark contrast to the standard,abstract,detachedexamplesnormallyprovidedinstatisticsbooks. I am most fortunate to have known these three authors for a few years now. Ihaveworkedwiththemallonmanyprojectsandhavehadthegoodfortunetosit for many hours, discussing all manner of things with them. They have produced abookthatwillnotonlyeducateyoubutalsogiveyouagoodread. BonAppe´tit! DebraBekerian Acknowledgments Wewouldliketotaketheopportunitytoexpressourgratitudetothemanypeople whohavehelpedmakethisbookpossible. Wewouldliketothankallofourfamily,friends,andlovedoneswhopatiently supportedusasweworkedonthisbook.Theirloveandsupporthelpedustomake thispossibleandforthatweareforevergrateful. We would also like to thank Kristin Rodgers, MLIS, Collects Curator, The MedialHeritageCenteroftheHealthSciencesLibraryattheOhioStateUniversity, Columbus,Ohio,forassistancewithstatisticaltablesandpermissionsandDanBell, Ph.D., Associate Professor of Mathematics, School of Arts and Sciences, Tiffin University,Tiffin,Ohio,foradviceandsupport. Finally,wewouldliketothankMarcStrauss,HannahBracken,andtheeditorial staffatSpringerScienceandBusinessMediafortheirguidanceandexpertise.This bookwouldnothavebeenpossiblewithouttheirbeliefinourwork. ix