Module 2: Data Preparation

Purpose

SmartPLS results are only as good as the dataset. This module teaches the data preparation steps needed before importing into SmartPLS 4.

Learning Objectives

After this module, learners should be able to:

• Prepare a SmartPLS-ready CSV file.
• Create a codebook for variables and indicators.
• Identify missing values and incorrect coding.
• Decide whether variables are metric, binary, or indicators of constructs.

SmartPLS Data Checklist

Before importing:

• First row contains variable names.
• Variable names are short and clear.
• No duplicate column names.
• Numeric variables use consistent decimal notation.
• Missing values are coded consistently.
• Binary variables are coded as 0/1.
• Reverse-coded items are corrected before analysis.
• Each row represents one respondent or case.

Variable Naming

Use names like:

PU1, PU2, PU3
TR1, TR2, TR3
SQ1, SQ2, SQ3
SAT
AGE
EXP

Avoid names with spaces, symbols, or very long descriptions.

Codebook

A codebook should include:

• Variable name
• Full label
• Construct
• Measurement scale
• Coding direction
• Missing-value treatment

Use Codebook Template.

Missing Data

Document:

• Number of missing values per variable.
• Whether missingness is random or systematic.
• Whether you used deletion or imputation.
• Whether SmartPLS case-wise deletion changes the sample size.

Reverse Coding

Reverse-coded survey items should be corrected before import.

For a 1 to 5 scale:

reversed = 6 - original

Practice

1. Open datasets/smartpls_regression_training_data.csv.
2. Identify the dependent variable for observed-variable regression.
3. Identify the predictor variables.
4. Identify the indicator variables for PLS-SEM.
5. Complete one row of the codebook template.

Lab

Complete Lab 1: Import dataset and create project.

Next Module

Continue to Module 3: Building a Regression Model.