Partial least squares path modeling: Time for some serious second thoughts

Journal of Operations Management - Tập 47 - Trang 9-27 - 2016
Mikko Rönkkö1, Cameron N. McIntosh2, John Antonakis3, Jeffrey R. Edwards4
1Aalto University School of Science, PO Box 15500, FI-00076, Aalto, Finland
2Public Safety Canada, 340 Laurier Avenue West, Ottawa, Ontario, K1A 0P8, Canada
3Faculty of Business and Economics, University of Lausanne, Internef #618, CH-1015, Lausanne-Dorigny, Switzerland
4Kenan-Flagler Business School, University of North Carolina at Chapel Hill, Campus Box 3490, McColl Building, Chapel Hill, NC, 27599-3490, USA

Tóm tắt

AbstractPartial least squares (PLS) path modeling is increasingly being promoted as a technique of choice for various analysis scenarios, despite the serious shortcomings of the method. The current lack of methodological justification for PLS prompted the editors of this journal to declare that research using this technique is likely to be desk‐rejected (Guide and Ketokivi, 2015). To provide clarification on the inappropriateness of PLS for applied research, we provide a non‐technical review and empirical demonstration of its inherent, intractable problems. We show that although the PLS technique is promoted as a structural equation modeling (SEM) technique, it is simply regression with scale scores and thus has very limited capabilities to handle the wide array of problems for which applied researchers use SEM. To that end, we explain why the use of PLS weights and many rules of thumb that are commonly employed with PLS are unjustifiable, followed by addressing why the touted advantages of the method are simply untenable.

Tài liệu tham khảo

10.1016/j.jom.2013.07.015 10.1287/isre.2014.0545 10.1145/2544415.2544417 10.1016/S0272‐6963(98)00018‐7 10.1016/j.leaqua.2010.10.010 10.2307/258555 10.1080/15366367.2015.1016349 10.1002/9781119970583 10.1016/j.lrp.2014.02.005 10.1007/978-0-387-45528-0 10.1177/1094428106294734 10.1002/9781118619179 10.1007/BF02296961 10.1080/10705510903008253 10.1037/met0000056 10.1177/004912417600500101 10.1016/j.jbusres.2012.08.006 10.1037/h0048255 Chin W.W., 1998, Modern Methods for Business Research, 295 10.1007/978-3-540-32827-8_29 10.1287/isre.14.2.189.16018 Chin W.W., 1999, Statistical Strategies for Small Sample Research, 307 10.2307/41703491 10.1177/1094428114555994 Chumney F., 2013, Structural Equation Models with Small Samples: a Comparative Study of Four Approaches 10.1037/0003-066X.45.12.1304 Cohen J., 2003, Applied Multiple Regression/correlation Analysis for the Behavioral Sciences 10.3102/10769986029003317 10.1017/CBO9780511802843 10.2307/23044046 10.1016/0304-4076(83)90094-5 10.1007/978-3-540-32827-8_2 10.1016/j.lrp.2014.02.004 T.K.Dijkstra.PLS & CB SEM a Weary and a Fresh Look at Presumed Antagonists (Keynote Address).2015 June Presented at the 2nd International Symposium on PLS Path Modeling Sevilla Spain. Retrieved from:http://www.researchgate.net/publication/277816598_PLS__CB_SEM_a_weary_and_a_fresh_look_at_presumed_antagonists_(keynote_address) 10.1016/j.csda.2014.07.008 10.25300/MISQ/2015/39.2.02 10.1007/s11336‐013‐9370‐0 10.1027/1614‐2241.3.3.100 10.1177/109442810142004 10.1177/1094428110378369 Efron B., 1983, A leisurely look at the bootstrap, the jackknife, and cross‐validation, Am. Statistician, 37, 36 10.1007/978-1-4899-4541-9 Enders C.K., 2010, Applied Missing Data Analysis 10.1007/978-3-540-32827-8_3 J.Evermann M.Tate.Testing models or fitting models? Identifying model misspecification in PLSICIS 2010 Proceedings. St. Louis MO.2010 Retrieved from:http://aisel.aisnet.org/icis2010_submissions/21 10.1016/j.jbusres.2016.03.050 10.2307/3151718 Fornell C., 1994, Advanced Methods of Marketing Research, 52 Fox J., 1997, Applied Regression Analysis, Linear Models, and Related Methods 10.1126/science.1255484 10.2307/23044042 Gefen D., 2005, A practical guide to factorial validity using PLS‐Graph: tutorial and annotated example, Commun. Assoc. Inf. Syst., 16, 91 10.2307/3172650 10.1287/isre.1070.0123 10.2307/41703490 D.L.Goodhue W.Lewis R.Thompson.PLS pluses and minuses in path estimation accuracyAMCIS 2015 Proceedings. San Juan Puerto Rico.2015 Retrieved from:http://aisel.aisnet.org/amcis2015/ISPhil/GeneralPresentations/3 Green S., 2010, Can specification searches be useful for hypothesis generation?, J. Mod. Appl. Stat. Methods, 9 10.1037//1082‐989X.6.1.67 10.1037/1082-989X.6.4.430 10.1016/S0272‐6963(15)00056‐X Hair J.F., 2014, A Primer on Partial Least Squares Structural Equations Modeling (PLS‐SEM) 10.2753/MTP1069‐6679190202 10.1016/j.lrp.2012.09.011 10.1007/s11747‐011‐0261‐6 10.1007/BF02289468 Hastie T., 2013, The Elements of Statistical Learning: Data Mining, Inference, and Prediction 10.1177/014662168500900204 10.1111/poms.12377 J.Henseler.Common Beliefs and Reality about PLS.2014 Retrieved from:http://managementink.wordpress.com/2014/05/23/common‐beliefs‐and‐reality‐about‐pls/ Henseler J., 2014, ADANCO (Version 1.0) Kleve 10.1177/1094428114526928 10.1108/IMDS‐09‐2015‐0382 10.1007/s11747-014-0403-8 10.1108/S1474-7979(2009)0000020014 10.1007/s00180‐012‐0317‐1 10.1145/2544415.2544418 Huang W., 2013, PLSe: Efficient Estimators and Tests for Partial Least Squares Hu L., 1999, Cutoff criteria for fit indexes in covariance structure analysis: conventional criteria versus new alternatives, Struct. Equ. Model. A Multidiscip. J., 6, 1, 10.1080/10705519909540118 10.1007/BF02295841 10.1016/j.jom.2003.12.001 Jöreskog K.G., 1982, Systems under Indirect Observation: Causality, Structure, Prediction. Amsterdam: North‐Holland 10.1016/j.jom.2004.05.003 10.1016/0272‐6963(94)90005‐1 10.1093/biomet/58.3.433 Kline R.B., 2011, Principles and Practice of Structural Equation Modeling Kock N., 2015, PLS‐based SEM algorithms: the good neighbor assumption, collinearity, and nonlinearity, Inf. Manag. Bus. Rev., 7, 113 Koopman J., 2014, More Statistical and Methodological Myths and Urban Legends, 224 10.1037/a0036635 10.1016/j.jom.2014.09.007 Krämer N., 2006, Analysis of High‐dimensional Data with Partial Least Squares and Boosting 10.1016/j.jbusres.2012.08.004 Lehmann E.L., 2005, Testing Statistical Hypotheses 10.2307/25148781 10.1016/j.ijresmar.2011.03.006 MacCallum R.C., 2007, Factor Analysis at 100: Historical Developments and Future Directions, 153 Marcoulides G.A., 2012, Handbook of Structural Equation Modeling, 690 10.2307/25148727 10.4324/9780203501207 10.1016/j.jom.2007.01.021 10.1207/s15327906mbr3102_5 10.3102/1076998607305835 10.1177/1094428114529165 10.1037/0003‐066X.30.10.955 10.1177/0013164496056001004 10.1016/j.jom.2012.03.001 10.1016/j.jom.2012.06.002 10.1016/j.jom.2012.11.002 10.1016/j.ijresmar.2009.08.001 10.4324/9781315736013 10.1080/15366367.2015.1016343 10.1177/1094428109332834 10.1207/s15327906mbr3003_4 10.1016/j.lrp.2012.09.010 10.1007/s13162‐013‐0034‐0 Ringle C.M., 2014, SmartPLS 3 10.1080/00207540902991682 Rönkkö M., 2014, Methodological Myths in Management Research: Essays on Partial Least Squares and Formative Measurement 10.1177/1094428114525667 M.Rönkkö.Introduction to matrixpls.2016 Retrieved from:https://cran.r‐project.org/web/packages/matrixpls/vignettes/matrixpls‐intro.pdf M.Rönkkö.matrixpls: Matrix‐based Partial Least Squares Estimation (Version 1.0.0).2016 Retrieved from:https://cran.r‐project.org/web/packages/matrixpls/index.html 10.1177/1094428112474693 M.Rönkkö J.Evermann M.I.Aguirre‐Urreta.Estimating Formative Measurement Models in IS Research: Analysis of the Past and Recommendations for the Future.2016 Unpublished Working Paper. Retrieved from:http://urn.fi/URN:NBN:fi:aalto‐201605031907 M.Rönkkö C.N.McIntosh M.I.Aguirre‐Urreta.Improvements to PLSc: Remaining Problems and Simple Solutions.2016 Unpublished Working Paper. Retrieved from:http://urn.fi/URN:NBN:fi:aalto‐201603051463 10.1016/j.paid.2015.07.019 M.Rönkkö J.Ylitalo.Construct validity in partial least squares path modelingICIS 2010 Proceedings. St. Louis MO.2010 Retrieved from:http://aisel.aisnet.org/icis2010_submissions/155 10.1016/j.jom.2009.03.001 10.1016/j.jom.2014.01.002 10.1108/S1474‐7979(2011)0000022012 10.1016/j.lrp.2014.02.008 Schroeder R.G., 2001, High Performance Manufacturing: Global Perspectives 10.1177/1094428115598239 10.1016/j.jom.2005.05.001 10.1214/10‐STS330 10.2307/23042796 10.1016/j.jbusres.2016.03.049 Singleton R.A.J., 2009, Approaches to Social Research 10.1007/BF02296196 10.1177/1059601108329198 10.2307/1412159 10.2307/2393788 10.1007/978-3-540-32827-8_32 10.1038/141246a0 10.1177/1094428105285506 10.1108/S1479-8387(2011)0000006009 10.1016/j.jom.2011.10.001 10.1016/0272‐6963(93)90034‐M 10.1007/978-3-540-32827-8_1 10.1007/s11747‐015‐0455‐4 10.1016/S0272‐6963(98)00019‐9 Westland J.C., 2015, Structural Equation Models, 10.1007/978-3-319-16507-3 Widaman K.F., 2007, Factor Analysis at 100: Historical Developments and Future Directions, 177 Wold H., 1966, Research Papers in Statistics: Festschrift for J. Neyman, 411 H.Wold.Factors Influencing the Outcome of Economic Sanctions: an Application of Soft Modeling.1980 Presented at the Fourth World Congress of Econometric Society Aix‐en‐Provence France H.Wold.Soft modeling: the basic design and some extensions.K.G.Jo¨reskog S.WoldSystems under Indirect Observation: Causality Structure Prediction.1982;1–54(Amsterdam: North‐Holland) 10.1007/BF02888567 10.1007/978-94-009-5079-5_11 Wooldridge J.M., 2009, Introductory Econometrics: a Modern Approach Yung Y.‐F., 1996, Advanced Structural Equation Modeling: Issues and Techniques, 195
Thông tin
Thông tin xuất bản

Journal of Operations Management

Tập 47

9-27

Thông tin tác giả