#SmartPLS4

The session focuses on how to interpret the measurement model output to check Reliability and Construct Validity.
Measurement Model assessment is the first step in the structural equation modelling. As part of Measurement Model Assessment the Construct Reliability is evaluated through Cronbach'a Alpha and Composite Reliability. Construct validity is checked through Convergent Validity and Discriminant Validity (Fornell & Larcker Criterion, Cross Loadings, and Heterotrait-Monotrait Ratio).
Next Session (How to Report Measurement Model Results):
• Video
How to Check Reliability in #SmartPLS4: • #SmartPLS4 Series 7.2 ...
Basic Concept of Convergent and Discriminant Validity: • #SmartPLS4 Series 6 - ...
How to Check Convergent Validity: • #SmartPLS4 Series 8 - ...
How to Check Discriminant Validity: • #SmartPLS4 Series 9 - ...
About:
Research With Fawad is a Research initiative to help young and early career research learn the research tools and techniques. It is an initiative to share research knowledge and help researchers learn to research the easy way.
Learn more about Research With Fawad at www.researchwi....
#SmartPLS4 #MeasurementModel #DiscriminantValidity #ConstructValidity #SmartPLS #plssem #sem #DiscriminantValidity #Problem_DiscriminantValidity

[@Alhamzah_F_Abbas]

Brilliant work

[@researchwithfawad]

Thanks. I am glad you liked it.

[@lina.r11]

Dear Sir. Thank you for the explanation! May I know whether we should use all three types of measurement of discriminant validity or can we choose one of them for the measurement model evaluation e.g. cross loading only

[@researchwithfawad]

Pleasure. I am glad you liked it. If you report only Cross-loadings, you may be asked to report HTMT.

[@lina.r11]

Dear Sir, thank you. Yes I learn a lot from your video. If cross loading reveals discriminant validity but the HTMT has not established the discriminant validity, so should we delete the items?

[@researchwithfawad]

If deleting doesn't affect how you have conceptualised the construct, you can.

[@dr.syedfaisal8830]

Mashallah

[@researchwithfawad]

جزاک اللہ خیرا

[@mastersilvaworld]

Dear Prof, is it ok to have negative values in the F-L table?

[@researchwithfawad]

Yes. The relationships can be negative

[@taibasahaf7774]

Thank you Sir for the knowledge that you impart to us. I have a question. In the video on construct reliability, you had mentioned that reliability values above 0.90 are not desirable. But here we got such values. What should one do when you get reliability values above 0.90 or as you pointed out in the previous videos, 0.95?

[@researchwithfawad]

Thanks. I am glad you liked it. Check for item to total correlation plus the correlation of items with each other in a particular construct. Too high correlation means two items are measuring the same concept. One of them can be removed and reliability assessed again.

[@taibasahaf7774]

@@researchwithfawad Jaza ka Allahu Khairah. Will do that!

[@denizhancay4901]

Hey Prof., thanks for your kindly explanation. What if we have a value higher than 1.000 at the HTMT?

[@researchwithfawad]

Thanks for watching. Then, there is a severe issue of discriminant validity.

[@bushrahal-fakeh771]

Hello Sir, please why you did not add a construct circle for "internal marketing" and "internal service quality" to connect with their dimensions?

[@researchwithfawad]

Because i am assessing reliability and validity of the lower order constructs. The reliability and validity of the lower order constructs is assessed first then we move to validity of higher order constructs.

[@bushrahal-fakeh771]

@@researchwithfawad Noted. Thank you your reply.

[@kamransaeed8755]

Sir, What to do if the p-value is greater than 0.05? How to fix this issue?

[@researchwithfawad]

P value greater than 0.05 is not an issue. It means your Hypothesis is not supported. You will need to discuss why the results are insignificant.