Search found 7 matches

by erikmonness
Tue Nov 01, 2016 1:11 pm
Forum: SmartPLS 3 - Requested Features
Topic: Data manipulations
Replies: 1
Views: 32824

Data manipulations

1) The smartPLS data file is fixed. Often one need to calculate/add some new indicators, make a transformation etc. Now one has to do this outside smartPLS and create a new project. This could be solved by having a data editor, or better detach data from calculations, like in most programs like SPSS...
by erikmonness
Tue Oct 11, 2016 11:07 am
Forum: Method and application
Topic: A formative construct within a cause chain
Replies: 1
Views: 2771

A formative construct within a cause chain

Consider the following part of a larger model. (ReflectiveX1, reflectiveX2…) --> (A formative construct) --> (ReflectiveY1…) Question: Can a formative construct be caused by something except its regressors? Do it make sense? Or has a formative construct have to be a leftmost construct in any chain?
by erikmonness
Mon Apr 25, 2016 11:13 am
Forum: SmartPLS 3 - Requested Features
Topic: Add variables, change data
Replies: 2
Views: 19586

Add variables, change data

Adding new variables to a smartPLS project need for a total redesign from scratch. In addition, when a good model is to be applied to a new data set, one has to redesign from scratch. . Something like “add variable” would be nice. The best thing could be that smartPLS used a EXCEL file as data sourc...
by erikmonness
Mon Aug 24, 2015 8:36 am
Forum: Method and application
Topic: Multi-Group Analysis
Replies: 1
Views: 2625

Multi-Group Analysis

There is a description that one can do Multi-Group Analysis. However, no “how”. The window that opens tells me to “select group”, and Groups A and groups B. But I find no option ta actually do so, like selecting a grouping variable. There are just red rectangles with a warning to “please select at l...
by erikmonness
Thu Nov 22, 2012 10:11 am
Forum: SmartPLS 2 - FAQ
Topic: PLS vs. sem-PLS
Replies: 0
Views: 2216

PLS vs. sem-PLS

Traditional partial least square finds directions in X-space and Y-space that have the highest covariance but also the highest variances. Using programs like jmp or SYSTAT the dimensionality chosen is the same in both spaces. How does that compare to the sem-PLS approach where one makes separate out...
by erikmonness
Tue Nov 20, 2012 3:01 pm
Forum: SmartPLS 2 - FAQ
Topic: Total regression coefficients
Replies: 1
Views: 2060

Total regression coefficients

I found out myself: Use original metric, multiplying path and outer loadings, gives total regression coefficients. Howver, the path plot only gives correlations whatever I do with preferences??
by erikmonness
Mon Nov 19, 2012 11:44 am
Forum: SmartPLS 2 - FAQ
Topic: Total regression coefficients
Replies: 1
Views: 2060

Total regression coefficients

When running PLS in jmp, SYSTAT etc, one gets the overall regression coefficients between every measured X and every measured Y, in SYSTAT including std’s. How can I get those from smartPLS