Measurement Issues in Cross-National Research
- PDF / 508,176 Bytes
- 23 Pages / 444 x 771 pts Page_size
- 56 Downloads / 210 Views
Abstract. Substantive inferences from cross-national studies have importantimplicationsfor theory (e.g., because they reveal insights into generalizabilityand boundaryconditions)and managerialpractice (e.g., becausethey offer guidelinesto MNC managers).However,few empiricalstudiesattendto measurement issuesinvolvedin cross-national research,and still fewer recognizethe risk of inferentialerrors that are likely to occur by overlookingmeasurementissues. We discuss fourmeasurement issues,namely(1) standardizedversusunstandardized coefficients,(2) the impactof measurement errorandunequalreliability, (3) the overallerrorrate and simultaneousanalysis and (4) construct equivalence.Using illustrativeexampleswe demonstratethe natureof each of these problems,the likely impactthey can haveon substantive conclusions,and approachesfor tacklingthese problems.Additionally, we reanalyze a recently publishedthree-nationstudy by Dubinsky, Michaels,Kotabe,Lim and Moon 119921to clarify these measurement concerns,highlighta methodologicalapproach,and delineatethe extent and severityof inferentialerrors.Our reanalysisshows that the interactive effects of these measurementissues are pervasive,complexand We close with implicationsfor cross-nationalresearchin unpredictable. general. We discussfour measurementissues in cross-nationalresearchand show that inattentionto these issues is problematicand, in many instances,results in serious inferentialand substantiveerrors.These issues include (1) the use of standardizedversus unstandardizedcoefficients,(2) measurementerror and unequalreliability,(3) overallerrorrate and (4) constructequivalence.While some of these issues have been discussedin the literature(JIBS, Fall 1983), our discussioncontributesto the literaturein three ways. First, none of the previous studies focus exclusivelyon measurementissues in cross-national research. Because these issues influence substantive inferences,a detailed, *Jagdip Singh is Associate Professor, Division of Marketing, Weatherhead School of Management, Case Western Reserve University, Cleveland, Ohio. This research was partially accomplished when the author was visiting fellow at the Timbergen Institute, School of Economics, Erasmus University, Rotterdam, The Netherlands. The author thanks three anonymousJIBS reviewersfor their helpfulcommentsthroughoutthe review process.In addition,the authoris indebtedto Roy Howellfor his generouscommentsand encouragement. Additionalstatisticaldetailsconcerningthe unrestrictedand restrictedmodels estimatedand reportedin this paperareavailablefromthe author. Received:January1994;Revised:August& December1994;Accepted:December1994 597
Palgrave Macmillan Journals is collaborating with JSTOR to digitize, preserve, and extend access to Journal of International Business Studies ® www.jstor.org
JOURNAL OF INTERNATIONAL BUSINESS STUDIES, THIRD QUARTER 1995
598
focused discussion of measurement problems is likely to be useful for researchers.Second, we discuss the measurement issues from an empirical and analytical standpoint.
Data Loading...
