SmartPLS

SmartPLS

Logo icon
Original author(s) Christian M. Ringle, Sven Wende, Jan-Michael Becker
Developer(s) SmartPLS GmbH
Initial release 2005 (2005)
Stable release
SmarPLS 3.2.6 / November 13, 2016 (2016-11-13)
Operating system Windows & Mac
Platform Java
Available in English
Type Statistical analysis, Data Collection, Structural equation modeling, Multivariate analysis
License Proprietary software
Website www.smartpls.com

SmartPLS is a software with graphical user interface for variance-based structural equation modeling (SEM) using the partial least squares (PLS) method.[1][2][3] The software can be used in empirical research to analyse collected data (e.g. from surveys) and test hypothesized relationships. Since it is programmed in Java, it can be executed and run on Windows and macOS operating systems.

Downloads

Data

SmartPLS uses raw data. The data import uses the text (.txt) and comma separated values (.csv) file format. The columns of the data file represent the variables; the rows represent observations and responses. Only the first row of the data matrix, which becomes the header in SmartPLS, can contain text. Otherwise, only numbers are allowed. Empty cells (e.g., if a missing value occurs) are not allowed. A number that is not used otherwise in the dataset (e.g., -999,999) must be used to fill empty cells. After data import, the user can indicate that this number represents a missing value. SmartPLS automatically identifies the data and number format (e.g., Europe or the USA). If a problem occurs, the user can later determine file format specifications and the number format in the software. In addition to the variables used as indicators for PLS path modeling, the data matrix can contain a weighting vector of the observations and responses (e.g., for carrying out weighted PLS path modeling analyses) and grouping variables (e.g., for conducting a PLS multigroup analysis, PLS-MGA).

Graphical User Interface

The graphical interface allows users to create a PLS path model. Circles represent latent variables and rectangular their indicators. The indicators stem from the variables of the imported data set. Drag and drop allows assigning indicators to a latent variable and, thereby, to establish the measurement model. In the structural model, an arrow drawing option allows the user to connect the latent variables with each other.

Algorithms

The primary basic partial least squares algorithm (Wold 1982, Lohmöller 1989) is the primary algorithm used in SmartPLS. The following list provides an overview of implemented algorithms and analytical options:

Blindfolding is a sample re-use technique that calculates the Stone-Geisser’s Q² value [4][5] as a criterion of predictive relevance.
Blindfolding is only applied to constructs with a reflective measurement model specification.
A technique to test for differences between identical models estimated for different groups.
A statistical technique which allows for empirical testing the measurement model setup.
A latent class approach which allows identifying and treating unobserved heterogeneity in path models.

Documentation

In-built Documentation. For every algorithm, an in-built documentation gives a short explanation. It also provides information regarding the parameter setting of the algorithms. References to key articles allow the user to access more detailed information about the specific algorithms and analysis options.

Results Report. For every analysis, SmartPLS offers a results report, which contains information about the model and data used for the analysis, key results, quality criteria (if relevant for a certain algorithm) and graphical results presentation (if applicable). The default offers a results representation within the SmartPLS software. In addition, SmartPLS allows exporting the results as a Microsoft Excel file, an HTML file, and in a file format that can be used by the statistical software R.

References

  1. Ringle, Christian M.; Wende, Sven; Will, Alexander (2005). SmartPLS 2. Hamburg: SmartPLS.
  2. Ringle, Christian M.; Wende, Sven; Becker, Jan-Michael (2015). SmartPLS 3. Boenningstedt: SmartPLS.
  3. Hair, Joseph F.; Hult, G. Tomas M.; Ringle, Christian M.; Sarstedt, Marko (2014). A Primer on Partial Least Squares Structural Equation Modeling (PLS-SEM). Thousand Oaks, CA: Sage.
  4. Stone, M.: Cross-Validatory Choice and Assessment of Statistical Predictions. In: Journal of the Royal Statistical Society. 36, 1974, S. 111–147.
  5. Geisser, S.: A predictive approach to the random effect model. In: Biometrika. Jg. 61, 1974, S. 101 ff..
This article is issued from Wikipedia - version of the 12/5/2016. The text is available under the Creative Commons Attribution/Share Alike but additional terms may apply for the media files.