SmartPLS 4 Tutorial Notes

Introduction

• How to start a new project with SmartPLS 4
• Navigate to top left and click on New Project
• Alternatively, go to File > New Project

Creating a New Project

• Click on the New Project icon
• Name the project (e.g., YouTube Tutorials) and hit Create
• New project appears on the left sidebar

Importing Data

• Click on Import Data File to explore and select a dataset
• Supports various file types:
  • .csv
  • .txt
  • .sav (SPSS)
  • .xls (Excel)
• Select the data file (e.g., SPSS file) and open it
• SmartPLS 4 performs a careful screening of the data

Variable Scaling

• Assign types of scales to variables:
  • ID: Metric
  • Q1, Q2: Ordinal
  • Age: Metric
  • Gender: Binary
  • Marital Status: Categorical
  • People in Household: Metric
  • Years in Job: Metric
  • Customer Interactions: Metric
  • Years in Firm: Metric
  • Education: Categorical
  • Income: Ordinal
  • Job Category: Binary
• Warning: 22 missing values found
  • Can specify a missing marker if applicable

Data View

• Basic descriptive statistics displayed
  • Type, missing values, mean, median, min/max, standard deviation
• Kramer's test for distributional normality
• Correlation matrix available by clicking on Indicator Correlations
• Option to edit setup and change values if needed
  • Return to workspace if no changes are made
• Ensure dataset status is green (i.e., ready to proceed)

Creating a Model

• Select model type: PLS SEM (Partial Least Squares Structural Equation Modeling)
• Name the model (e.g., YouTube Model)
• New interface for model design:
  • Selector arrow, latent variable tool, connector tool
• Start creating a basic model
  • Select variables (e.g., Satisfaction with Work, Burnout)
  • Name latent variables appropriately
• Connect variables to form relationships:
  • Ethical concerns → Burnout
  • Burnout → Satisfaction with Work
  • Ethical concerns → Satisfaction with Work

Saving and Calculating Results

• Navigate to Save and Calculate
• Options available: standard PLS SEM algorithm, consistent algorithms
• Settings for path or factor
• Choose result type (standardized, unstandardized, or mean-centered)
• Start calculation
  • Outputs standardized regression weights, R-squared values
  • Hovering over variables displays detailed information (e.g., R-squared, Cronbach’s alpha, composite reliability, average variance extracted)

Conclusion

• Subsequent videos will cover additional options for analysis in SmartPLS 4.