LIKERT SCALE DATA ANALYSIS METHODOLOGY
Generated on: 6/17/2025

OVERVIEW:
This document outlines the methodology used for generating and analyzing synthetic Likert scale data.

DATA GENERATION:
- Scale Type: 5-point Likert scale (1 = Strongly Disagree, 5 = Strongly Agree)
- Sample Size: 300 synthetic respondents
- Variables: 3 latent constructs
- Items per Variable: Kepuasan Kerja: 4 items, Keseimbangan Kehidupan Kerja: 4 items, Komitmen Organisasi: 4 items

RELIABILITY ANALYSIS:
- Cronbach's Alpha: Internal consistency reliability measure
  - Excellent: α ≥ 0.9
  - Good: 0.8 ≤ α < 0.9
  - Acceptable: 0.7 ≤ α < 0.8
  - Poor: α < 0.7

- Test-Retest Reliability: Stability over time
- Inter-Rater Reliability: Agreement between raters

VALIDITY ANALYSIS:
- Average Variance Extracted (AVE): Convergent validity measure calculated from factor loadings
  - Acceptable: AVE ≥ 0.5
- Convergent Validity: Items measuring same construct should correlate highly
- Discriminant Validity: Items measuring different constructs should correlate weakly (Fornell-Larcker criterion)

FACTOR ANALYSIS:
- Principal Component Analysis (PCA): Extracts underlying factors from item correlations
- Factor Loadings: Correlation between items and extracted factors
  - Strong: |loading| ≥ 0.7
  - Moderate: 0.5 ≤ |loading| < 0.7
  - Weak: |loading| < 0.5
- Kaiser-Meyer-Olkin (KMO): Sampling adequacy measure
  - Excellent: KMO ≥ 0.8
  - Good: 0.7 ≤ KMO < 0.8
  - Mediocre: 0.6 ≤ KMO < 0.7
  - Miserable: 0.5 ≤ KMO < 0.6
  - Unacceptable: KMO < 0.5
- Bartlett's Test: Tests sphericity assumption for factor analysis
- Communalities: Proportion of variance in each item explained by extracted factors
- Eigenvalues: Amount of variance explained by each factor
- Composite Reliability: Internal consistency measure using factor loadings

DESCRIPTIVE STATISTICS:
- Central Tendency: Mean, Median, Mode
- Variability: Standard Deviation, Variance
- Distribution Shape: Skewness, Kurtosis

NORMALITY TESTING:
- Shapiro-Wilk Test: Tests null hypothesis that data is normally distributed
- Significance Level: α = 0.05
- Skewness: Measure of asymmetry
  - Normal: -1 < skewness < 1
- Kurtosis: Measure of tail heaviness
  - Normal: -2 < kurtosis < 2

HYPOTHESIS TESTING:
H1: Kepuasan kerja berpengaruh positif terhadap komitmen organisasi (positive relationship expected)
H2: Keseimbangan kehidupan kerja berpengaruh positif terhadap kepuasan kerja (positive relationship expected)
H3: Keseimbangan kehidupan kerja berpengaruh positif terhadap komitmen organisasi (positive relationship expected)

STATISTICAL ANALYSES:
- Pearson Correlation: Measures linear relationships
- t-Tests: Compare means between groups
- ANOVA: Compare means across multiple groups
- Effect Size: Cohen's conventions (small: 0.2, medium: 0.5, large: 0.8)

LIMITATIONS:
- Data is synthetically generated for demonstration purposes
- Results should not be interpreted as real research findings
- Missing data handling: Listwise deletion applied
- Assumes interval-level measurement for Likert scales

SOFTWARE INFORMATION:
- Platform: Likert Data Analysis Tool
- Statistical Methods: JavaScript-based implementations
- Export Formats: Excel (.xlsx), PDF, PNG images
