STATISTICAL SOFTWARE
SmartPLS Analysis Service
PLS-SEM Models With Reproducible Inference
Our SmartPLS analysis service carries out PLS-SEM in SmartPLS, working from your data and project files, with measurement model assessment, bootstrapped path estimates and the resampling settings documented where agreed so that each result can be rerun and reported.
No Commitment Required
Strictly Confidential
In-House PhD-Level Statisticians
Fixed Project Fee
Same-Day Response
What Is SmartPLS Used For in Research?
SmartPLS is commercial software for partial least squares structural equation modeling (PLS-SEM), a composite-based method used for models with many constructs, formative indicators or a predictive aim. Researchers choose it for a graphical model editor with bootstrapping and the assessment of measurement and structural models built in. PLS-SEM is not covariance-based SEM, and its quality criteria differ: reliability, convergent and discriminant validity, collinearity, bootstrapped path coefficients and out-of-sample prediction. OnlineSPSS, a specialist statistical consultancy for academic, clinical and professional research, provides a SmartPLS analysis service in which an in-house PhD-level statistician carries out the agreed analysis in SmartPLS, working from your files and recording the settings so that you can rerun it.
WHEN TO ENGAGE
Four Situations Where Researchers Hand Us a SmartPLS Model
Projects arrive with a drawn model, a validity question from a reviewer, a doubt about construct type or an existing project that has to be reported.
Model Drawn, Assessment Pending Your constructs, indicators and hypotheses are defined, and the PLS path model has to be estimated, assessed in two stages and reported with bootstrapped intervals.
Discriminant Validity Questioned Only Fornell-Larcker or cross-loadings were checked, or a reviewer asks for HTMT, and the measurement model has to be reassessed before the structural results are trusted.
Reflective or Formative in Doubt It is unclear whether a construct is reflective or formative, which changes how its indicators are assessed and what counts as a problem in the results.
Project to Check and Report An existing SmartPLS project needs its settings verified by a SmartPLS statistician and its results written up in the form your target journal expects.
ANALYSES IN SMARTPLS
Analyses Carried Out in SmartPLS for Your Survey Data
SmartPLS PLS-SEM projects are built from a few recurring analyses, and each card links to the method page where the choice of approach is explained.
PLS-SEM (Partial Least Squares SEM)
The measurement model is assessed first, with loadings, reliability, convergent and discriminant validity, and only then the structural model, with collinearity, paths, R-squared and effect sizes.
Mediation Analysis
Indirect effects are bootstrapped and reported with confidence intervals, then read together with the direct effect, so the type of mediation is described from the estimates.
Moderation Analysis
The interaction term is built by a stated approach, and the effect is probed at meaningful levels of the moderator, not judged from the interaction coefficient alone.
Bootstrapping
SmartPLS offers percentile and bias-corrected and accelerated (BCa) intervals, and the interval type and one- or two-tailed testing are fixed in the analysis plan before any result is seen.
Reliability Analysis (Cronbach's Alpha)
Composite reliability and rho_A are reported beside Cronbach's alpha, because alpha assumes equal indicator loadings and tends to understate reliability when they differ.
Likert Scale Analysis
Items are screened for response patterns, ceiling effects and missing values before they enter the model as indicators of a construct.
COMMON SMARTPLS ISSUES
Where PLS-SEM Analyses Go Wrong Before the Paths
Most weaknesses in a PLS-SEM paper sit in the measurement model or in how inference was set up, so each check below is made before structural results are reported.
HTMT Beyond Fornell-Larcker Fornell-Larcker and cross-loadings detect weak discriminant validity poorly. HTMT is reported against a stated threshold, and SmartPLS HTMT bootstrapping adds an interval to the ratio.
Resamples and a Fixed Seed Too few bootstrap resamples make intervals unstable between runs. The number is set high, the seed is fixed and both are recorded, so that a rerun returns the same intervals.
The Legacy Global GoF Index The old global goodness-of-fit index is not accepted as proof of model quality. Measurement and structural models are assessed separately, with out-of-sample prediction where prediction is the aim.
Reflective or Formative Specification Construct type is decided from theory before estimation: reflective constructs are judged by loadings and reliability, formative ones by indicator weights and collinearity.
Sample Size Judgment A blanket rule such as ten times the largest number of paths is not relied upon. The sample is judged from the model's structure and the smallest effect of interest.
Collinearity and Weak Indicators Collinearity among indicators and predictors is examined with VIF values, and a weak indicator is removed only with a content reason, never solely to improve the numbers.
DELIVERABLES
What You Receive From a SmartPLS Analysis
The files follow a documented SmartPLS project, agreed in the scope and provided where applicable, so that your team can reopen the model and see how each value arose.
SmartPLS Project The project file with the model, the data and the algorithm settings, delivered as run, so that the analysis can be reopened and repeated.
Settings and Seed Record The algorithm settings, the number of resamples, the interval method and the seed are written out, so that the inference can be reproduced.
Measurement Model Tables Tables of loadings, reliability, convergent validity and HTMT with intervals, laid out as journals in your field expect them.
Structural Model Results Path coefficients with bootstrapped intervals, R-squared, effect sizes and collinearity values, with prediction results where they were agreed.
Model Diagram and Figures A clear diagram of the estimated model and publication-ready figures, with the sample size and the settings stated on each.
Reporting Text Statistical wording for the methods and results, matched to the output and stating what the estimates support, for you to check and incorporate.
OTHER PLATFORMS
When Another Platform May Be More Appropriate
SmartPLS suits composite-based models and prediction, and when the question calls for common-factor models or another team's standard, a different platform fits better.
ILLUSTRATIVE EXAMPLE
How a SmartPLS Analysis May Look in Practice
An illustration of how a survey-based model moves from a borderline validity result to reporting that states its limits.
ILLUSTRATIVE EXAMPLE · TOURISM DESTINATION LOYALTY RESEARCH · UNITED STATES
A business school team surveyed visitors to model how perceived destination image and satisfaction relate to revisit intention, with every construct measured by several Likert items.
The model was specified from theory before estimation, with all constructs reflective. The measurement assessment showed acceptable loadings and reliability, but the HTMT ratio between two image constructs was high, and its bootstrapped interval was not clearly below the threshold. The wording of the items was reviewed, the constructs were kept separate only after the interval and the content of the items had been weighed together, and the borderline result was reported as a limitation. The resamples were set high with a fixed seed so the structural intervals could be reproduced, and the project was saved with its settings.
Outcome: The team received the measurement tables with HTMT and its interval, the structural estimates with bootstrapped 95% confidence intervals, the SmartPLS project file and a record of the settings.
This example does not describe a real client or project. We never share details of actual client work, in line with the confidentiality we promise every researcher. This illustrative example is included to show the type, scope, and standard of work involved in this kind of analysis.
RESEARCH DISCIPLINES
Research Fields Where SmartPLS Is Common
SmartPLS is often used in business, management and marketing research, and the same measurement questions arise in the social and health sciences wherever survey constructs are modeled.
HOW WE ASSESS YOUR PROJECT
We Review Your Actual Research
Your assessment is based on a review of your actual research materials and analytical requirements, not a generic project category.
You Submit Your Project Details
Share whatever is available at the current stage, such as the research aim, protocol or proposal, dataset, variable list, manuscript or statistical requirements.
We Review Your Research and Define the Scope
Your statistically trained Research Coordinator reviews the materials to understand the study design, research questions, data structure, analytical requirements and expected deliverables. If clarification is needed, additional documents or information can be requested before the assessment is finalized.
Your Analysis Is Completed and Delivered
Once the scope and fixed project fee are agreed, the in-house statistician completes the work and the Research Coordinator remains the communication point through delivery, with the analysis, deliverables and timeline as agreed.
RESEARCHER VOICE
What Researchers Say
Feedback from a marketing researcher on keeping the work in SmartPLS, discriminant validity and project files that made later changes easier.
“I needed the work kept in SmartPLS because that is what our team uses internally. The main problem was discriminant validity being judged from older checks without looking properly at HTMT. OnlineSPSS corrected the setup and returned the project files with clear reporting, which made later changes much easier.”
Statistical Support Within Your Research OnlineSPSS provides the agreed methodological, analytical and statistical reporting support, including the SmartPLS files described in the scope. The researcher or research team retains the scientific question, substantive interpretation, conclusions and overall responsibility for the research.
FREQUENTLY ASKED QUESTIONS
Questions About SmartPLS Analysis for Research
Short answers to what researchers ask before sharing a SmartPLS project, from the analyses and predictive power to group comparison, existing projects, confidentiality and the assessment.
Which parts of a PLS-SEM study do you carry out in SmartPLS?
We carry out the analyses researchers most often run in SmartPLS: PLS-SEM measurement and structural model assessment, mediation, moderation, bootstrapped inference and, where agreed, out-of-sample prediction. The choice follows your theory and the way your constructs are measured, and it is confirmed during the assessment.
Can you assess the predictive power of my model in SmartPLS?
Yes, where prediction is part of your aim. PLSpredict compares the out-of-sample prediction errors of the model with a simple linear benchmark, so predictive power is judged on cases the model was not estimated on, not from R-squared alone. The choice between PLS-SEM and covariance-based SEM is explained on Structural Equation Modeling (SEM).
Can you compare groups, such as two countries, in my SmartPLS model?
Yes, where each group is large enough. Measurement invariance is established first with the MICOM procedure, and path coefficients are compared across groups with multigroup analysis only after that, because a path difference means little if the constructs were measured differently.
Can my existing SmartPLS project be corrected without rebuilding the model?
Yes. The project and the data are opened as they are, the model specification and the settings are checked, and only what needs to change is changed, with each change recorded. If the model has already been reported, Independent Statistical Review checks the HTMT results and bootstrap settings before resubmission.
How do you protect my research data and confidentiality?
Your survey data and SmartPLS projects are handled confidentially, and project information is not published. An NDA is available where required and can be arranged before sensitive materials are shared. Read more about confidentiality and data security.
What happens when I request an assessment, and how long does it take?
You share your research question, design and relevant materials, such as the dataset and any existing project, and a statistically trained Research Coordinator reviews the request and responds the same day. The analysis, deliverables, timeline and one fixed project fee are then confirmed before any work begins.
Still have a question? Find more answers in our full FAQ.
NEXT STEP
Discuss Your
SmartPLS Analysis With Us
Share the research question, the construct list and the data, and any project you have started. We confirm the model specification and the inference settings before any work begins.
No Commitment Required
Strictly Confidential
Fixed Project Fee
In-House PhD-Level Statisticians
Same-Day Response
Serving Researchers Since 2014