Example: marketing

PLS Path Modeling with R - Gaston Sanchez

PLS path Modelingwith work is licensed under aCreative Commons Attribution-NonCommercial-ShareAlike UnportedLicense (CC BY-NC-SA ) short: Gaston Sanchez retains the Copyright but you are free to reproduce, reblog, remix and modify thecontent only under the same license to this one. You may not use this work for commercial purposesbut permission to use this material in nonprofit teaching is still granted, provided the authorshipand licensing information here is , G. (2013)PLS path Modeling with RTrowchez Editions. Berkeley, Jessicafor her manifest love,her latent support,her outer energy,and her inner wish my college teachers would have taught many more things than the onesthey did. Preferably, I would have liked to be taught more computational coursesand programming classes. They taught me the basics but they missed one keylesson that I learned years later with a blend of great disillusionment and enrichingexperience:the important Modeling should take place in the mind of the analyst,not in the computer.

Citation Sanchez, G. (2013) PLS Path Modeling with R Trowchez Editions. Berkeley, 2013. http://www.gastonsanchez.com/PLS PathModelingwithR.pdf i

Tags:

  With, Modeling, Path, Pls path modeling with r, Pls pathmodelingwithr, Pathmodelingwithr

Information

Domain:

Source:

Link to this page:

Please notify us if you found a problem with this document:

Other abuse

Transcription of PLS Path Modeling with R - Gaston Sanchez

1 PLS path Modelingwith work is licensed under aCreative Commons Attribution-NonCommercial-ShareAlike UnportedLicense (CC BY-NC-SA ) short: Gaston Sanchez retains the Copyright but you are free to reproduce, reblog, remix and modify thecontent only under the same license to this one. You may not use this work for commercial purposesbut permission to use this material in nonprofit teaching is still granted, provided the authorshipand licensing information here is , G. (2013)PLS path Modeling with RTrowchez Editions. Berkeley, Jessicafor her manifest love,her latent support,her outer energy,and her inner wish my college teachers would have taught many more things than the onesthey did. Preferably, I would have liked to be taught more computational coursesand programming classes. They taught me the basics but they missed one keylesson that I learned years later with a blend of great disillusionment and enrichingexperience:the important Modeling should take place in the mind of the analyst,not in the computer.

2 Computers and software are indispensable but they can tmake sense of data; only people who have developed the necessary data analysisskills can. Whether it is PLS path Modeling or any other type of Modeling ,always remember this little-big ModelContentsPreface .. vii1 Introduction .. PLS path Modeling .. R packageplspm.. R? .. About Partial Least Squares .. tale of PLS .. and path -modelists .. Structure of the book .. R andplspm.. Extra Details .. Reading List ..132 Getting Started with PLS-PM .. Case Study: Index of Success .. latent and manifest variables .. Success, Attack and Defense .. Model .. and Outer models .. Applying PLS-PM with the R packageplspm.. the ingredients forplspm().. ().. results .. Reading List ..313 Setting the PLS-PM Framework .. PLS-PM Frame of Mind .. PLS path Modeling Specs.

3 Formal Model:What we want.. Structural Model .. Measurement Model .. Weight Relations .. Operative Model:How to get what we want.. 1: Iterative Process .. 2: path Coefficients .. 3: Loadings .. up .. Reading List ..474 Interpreting PLS-PM Results .. Reminder of Our Success Index Model .. withplspm.. functionplspm().. Handling PLS-PM results .. Measurement Model Assessment:Reflective Indicators .. for one and one for all .. of indicators .. and Communalities .. Measurement Model Assessment:Formative Indicators .. Structural Model Assessment .. of determinationR2.. Validation .. Reading List ..725 Running a PLS-PM Analysis .. Modeling Customer Satisfaction .. Model .. Study: Satisfaction in Education .. Preliminary Exploration .. at the data .. the forest, not just the trees.

4 PLS-PM Round 1 .. PLS-PM Round 2 .. PLS-PM Round 3 .. unidimensionality .. and Communalities .. Coefficients .. Regressions .. Model Summary .. Validation .. scores of latent variables .. Guidelines for a PLS-PM analysis .. Reading List .. 1136 Comparing Groups .. The million dollar question .. Group Comparison Approaches .. t-test .. Procedure .. Study: College GPA .. Analysis: GPA Model .. and Male Student Models .. Comparing Groups: Bootstrap t-test .. Comparing Groups: Permutation test .. Reading List .. 1337 Moderating Effects .. Moderator Variables .. Moderating effects in PLS-PM .. variables as constructs .. Study: Simplified Customer Satisfaction .. Product Indicator Approach .. Two-Stage path Modeling Approach.

5 Two-Stage Regression Approach .. Categorical Variable Approach .. Reading List .. 1558 PLS path Models with Higher-Order Constructs .. : How to model higher-orderconstructs? .. types of higher-order constructs .. higher-order constructs in PLS-PM .. Study: NFL data .. Repeated Indicators:The Poor Man s Approach.. Indicators example .. Two-Step Option:The Patch Approach.. example .. Hybrid Option:The Give Away Approach.. Approach example .. Wrapping Up .. Reading List .. 1759 Detecting Classes with REBUS-PLS .. Diversity in Data .. Classes .. What is REBUS? .. REBUS Approach .. Study: Success Index Model .. PLS path Model .. Applying REBUS .. Models .. of REBUS models .. Comments .. Reading List.

6 201A Historical Overview of PLS-PM .. 2051 Preliminaries .. 2052 Herman Wold .. Econometrician .. Least Squares Affair .. Philosopher of Science .. 2133 Pre-PLS Era .. 2144 Soft Modeling .. 2175 PLS-PM Related Timeline .. 2186 Reading List .. 221CC BY-NC-SA Gaston SanchezPLS path Modeling with RPrefaceOn a sunny September morning of 2005, Tomas Aluja my PhD advisor and I werewalking around the campus of the Universitat Politecnica de Catalunya in Barcelona. So,how did you like the symposium? Tomas asked me. The week before we had hosted thePLS 05 International Symposium on PLS and Related Methods. It was my first academicconference ever and I had just been introduced to the PLS methods early that year. Ikind of liked it I responded. What was your favorite talk? , Tomas said. Well, I likedthis guy s presentation and also the case study in that woman s talk I answered vaguely.

7 Iwas part of the staff and my role during the symposium was to be in charge of one of theconference halls. In other words, I had to make sure that everything worked fine, the lights,the projector, the microphone, the speakers banners. I barely had a chance to sit down andlisten carefully to the talks. You can t blame me for not paying attention to the speakers;after all, I was checking that all the small details were in place (which I diligently did).Tomas paused, stared at me, and put all the cards on the table: Would you be interestedin working on PLS? I knew that a proposal like that from Tomas was a once-in-lifetimeopportunity that I couldn t reject. Yes, absolutely , I responded. I was not exactly sureif PLS was for me but I would figure that out later. The plan is to make an R library tocalculate PLS path models. Then, we need to work on segmentation issues.

8 You know, finda way to detect segments and compare path models .. Tomas continued describing thegeneral idea of the project, thinking out loud and mentioning some of the potential problemsthat I would have to deal with . When he finished, all I said was OK (I m not a verytalkative person). Oh boy, what an innocent grad student I was. To put that in perspectiveis like if you were just start learning to climb, and then your instructor says: Would yoube interested in Mount Everest? The plan is to climb mount Everest under the followingconditions: you need to climb it by its north face, during the winter season, in alpine style,without Sherpas, with no oxygen supply, and open a new route that nobody else has climbedbefore . Anyone in their right mind would thought twice about it and even then would havedoubted the feasibility of the project. But I m glad I didn t know in advance what it wouldtake to climb mount PLS.

9 I was going to climb it or die to July 2009. The seemingly utopian project my advisor proposed me, becamea reality. I ll save the story of what happened in those four years for when I write myixxmemoirs. The important thing for this story was what happened on the day of my dissertationdefense. That evening, after successfully defending my climbing of mount PLS, I attendedthe traditional dinner with Tomas, the members of the defense committee, and my invitedguests. At the end when it came the time to say goodbye, Professor Ludovic Lebart, amember of my defense committee, approached me and strongly encouraged me to write atextbook about Partial Least Squares path Modelingas soon as possible. Me? A book onPLS path Modeling ? No way! , I thought. At least not anytime soon. I smiled and told himthat I would think about it. Since I had just finished my PhD, I didn t have the necessarymotivation and required stamina to write another manuscript.

10 Besides, theHandbook ofPartial Least Squareswas already being edited. The world didn t need another book onPLS. Or so I has this strange way of proving that sometimes you are wrong. I kept maintaining theR package for PLS path Modeling and adding more features. Emails from all over the worldfilled my inbox with questions, inquiries, doubts, comments, suggestions, contributions, andof course, complaints about those random but nasty, pain-in-the-butt, bugs. People askedme how to apply PLS for a given problem; how to calculate such and such; what was themeaning of those displayed numbers; and so on. Moreover, Lebart s words were still in myhead. Maybe it wouldn t be such a bad idea to write a book on PLS path Modeling afterall , I used to say to myself. The more I remained involved with PLS either by maintainingthe R package, applying it for my research projects, answering emails and collaborating withother colleagues the more I realized there was something missing.


Related search queries