STATISTICS BASED ON SMARTPLS - ENGLISH FOR GENERAL PURPOSE (ANGELA VICTORIA - 2414190037)

Summary

In this video, Angela Victoria explains the evaluation of measurement and structural models using Smart PLS. She covers key aspects of measurement model evaluation, such as reliability and validity, including methods like outer loadings, Cronbach’s Alpha, and discriminant validity. The structural model evaluation focuses on testing direct and indirect effects between constructs, along with the use of bootstrapping to assess significance. The video provides an insightful overview of how to ensure that research instruments are both reliable and valid, and how relationships between constructs can be tested effectively in Smart PLS.

Takeaways

• 😀 Measurement model evaluation ensures that a research instrument accurately measures the intended concept.
• 😀 Reliability and validity are the two main aspects of measurement model evaluation.
• 😀 Outer loadings show how strongly each question relates to the concept being measured, with values above 0.7 being considered strong.
• 😀 Cronbach’s Alpha, Rho A, and Composite Reliability are used to measure internal consistency, with values above 0.7 being ideal.
• 😀 Average Variance Extracted (AVE) is used to measure how much information is captured by the questions, with values above 0.5 being considered good.
• 😀 Discriminant validity ensures that constructs are distinct from each other, and methods like the Fornell-Larcker Criterion and cross-loading are used to assess it.
• 😀 The Heterotrait-Monotrait Ratio (HTMT) is another method for assessing discriminant validity, with values below 0.85 or 0.9 indicating good validity.
• 😀 The structural model in Smart PLS evaluates if relationships between concepts are significant and important.
• 😀 Direct effects test if one concept directly affects another, with a p-value under 0.05 indicating significance.
• 😀 Indirect effects test if one concept affects another through a mediator, with a p-value under 0.05 indicating significance.
• 😀 Bootstrapping in Smart PLS resamples data to check if relationships between concepts are reliable, with a p-value under 0.05 indicating a significant effect.

Q & A

What is the main focus of the measurement model evaluation in Smart PLS?

The measurement model evaluation in Smart PLS focuses on ensuring that the research instrument, such as a questionnaire, accurately measures the concept it is intended to measure. This involves evaluating both reliability (the consistency of the results) and validity (whether the instrument truly measures the intended concept).

What are outer loadings in the context of measurement model evaluation?

Outer loadings represent how strongly each question in the measurement model relates to the concept being measured. Values above 0.7 indicate strong relevance, while values below 0.5 or 0.7 suggest weak relevance.

What is the role of Cronbach’s Alpha in assessing reliability?

Cronbach’s Alpha is a measure of internal consistency within the measurement model. Values above 0.7 are considered ideal, indicating good reliability. This means the items or questions in the instrument are consistently measuring the same underlying construct.

How does rho A differ from Cronbach’s Alpha?

Rho A is a newer and more accurate alternative to Cronbach’s Alpha for assessing reliability in complex models. While both assess internal consistency, rho A is preferred for more advanced and intricate measurement models due to its ability to handle complex constructs more effectively.

What is the significance of Composite Reliability in measurement model evaluation?

Composite Reliability is similar to Cronbach’s Alpha but is considered better for complex models. It provides a more reliable measure of internal consistency by evaluating the construct as a whole. Values above 0.7 indicate good reliability.

What is Average Variance Extracted (AVE) and how is it used in validity assessment?

Average Variance Extracted (AVE) measures how much of the total variance in a set of items is explained by the construct. It is used to assess convergent validity, and values above 0.5 are considered good, meaning the construct captures sufficient information from its indicators.

What is discriminant validity and why is it important in Smart PLS?

Discriminant validity ensures that constructs in the measurement model are distinct from each other and do not overlap. It is important because it confirms that the constructs represent different concepts, ensuring the accuracy of the research.

What methods can be used to test discriminant validity in Smart PLS?

Three methods can be used to test discriminant validity: 1) Fornell-Larcker Criterion, where the square root of AVE for each construct should be greater than its correlation with other constructs. 2) Cross-loadings, where a question should load higher on its intended construct than on others. 3) HTMT (Heterotrait-Monotrait Ratio), where values below 0.85 or 0.9 suggest good discriminant validity.

What is the purpose of the structural model in Smart PLS?

The structural model in Smart PLS is used to test the relationships between concepts, examining whether these relationships are significant. It evaluates direct effects, indirect effects (mediated relationships), and overall model fit to ensure that the hypothesized connections hold up empirically.

What is bootstrapping in Smart PLS and how is it used?

Bootstrapping is a resampling technique used in Smart PLS to check the reliability of the relationships in the structural model. By repeatedly resampling the data, it estimates standard errors and significance levels. If the P-value is below 0.05, the relationship is considered statistically significant.