PLS SEM Second Order menggunakan smartPLS Pada Indikator Refleksif

Summary

In this video, Tri Widya Pangestika explores second-order reflective indicators using SmartPLS, detailing the two-tier testing process involving latent variables. The tutorial illustrates how to structure a second-order model with first-order indicators, analyze validity through loading factors, and assess reliability via Cronbach's Alpha. Key concepts like convergent and discriminant validity are discussed, emphasizing statistical significance through p-values and t-statistics. By the end, viewers learn to interpret results effectively, including how various factors reflect overall characteristics, enhancing their understanding of advanced structural equation modeling.

Takeaways

• 📊 The second-order model in SmartPLS involves testing latent variables at two levels: first-order indicators and second-order latent variables.
• 🔍 Reflective indicators are characterized by a relationship from latent variables to their indicators, allowing for assessment of construct validity.
• 📈 A model's effectiveness can be visually represented, showing how first-order latent variables serve as indicators for a higher-order latent variable.
• 📝 Each first-order latent variable can have multiple indicators, and the relationship is established through factor loading analysis.
• 🛠 SmartPLS allows for easy importation of data and visualization of model diagrams, aiding in the clarity of complex relationships.
• ✅ Validity is determined through criteria such as loading factor values, with a threshold of above 0.7 indicating strong indicators.
• 🔒 Reliability of a model is assessed through Cronbach's Alpha and composite reliability, both needing to exceed 0.7 for confidence in the results.
• 📏 Discriminant validity can be checked using the Fornell-Larcker criterion, ensuring that the square root of AVE exceeds the correlations between constructs.
• 📊 The R-squared value indicates the explanatory power of the model, with values greater than 0.7 considered strong.
• 📉 Statistical significance is confirmed when the p-value is less than 0.05 or when the t-statistic exceeds 1.96, validating the relationships between variables.

Q & A

What is the focus of the video presented by Tri Widya Pangestika?

The video focuses on second-order sample modeling using reflective indicators in SmartPLS.

What does 'second-order' refer to in the context of latent variables?

Second-order refers to the testing conducted on two levels: first on latent variables with their indicators, and second on the relationships between first-order latent variables and the second-order latent variable.

How many first-order latent variables are illustrated in the video?

There are five first-order latent variables illustrated in the video.

What criteria are used to evaluate the goodness of fit for the model?

The model's goodness of fit is evaluated based on convergent validity, discriminant validity, and reliability.

What is the significance of a loading factor greater than 0.7?

A loading factor greater than 0.7 indicates that the indicators are valid and reflect the underlying latent variables well.

What is the role of Cronbach's Alpha in the analysis?

Cronbach's Alpha assesses the reliability of the measurement; a value above 0.7 indicates acceptable reliability.

How is discriminant validity assessed in the video?

Discriminant validity is assessed using the Fornell-Larcker criterion, which compares the square root of average variance extracted (AVE) with correlations between latent variables.

What does a p-value less than 0.05 indicate?

A p-value less than 0.05 indicates that the relationship between the variables is statistically significant.

What is the significance of the R-squared value in the model?

The R-squared value indicates the proportion of variance in the dependent variable that can be explained by the independent variables; values above 0.7 suggest a strong model.

Which first-order variable was found to have the highest contribution to the second-order latent variable?

The variable 'skill variety' had the highest contribution to the second-order latent variable, with a value of 0.963.