Single indicator for each latent variable

[baselabuali]

Dears

My model is a "pass" model only by using a single indicator for each latent variable. Accordingly, most of the measurement model (outer loadings) and structural model (Cronbach alpha, rho_A, composite reliability, AVE, VIF) are shown as "1.000". Is this OK??

Thanks

[jmbecker]

If you have no measurement model (only single indicators) there is nothing to calculate and the assumption for single indicators is perfect reliability (i.e., the construct is the same as the indicator). So this is ok.

However, the question is, whether using only single indicator constructs is in general ok. But that depends on your context and is an entirely different question.

[baselabuali]

Thank you Dr. Jan

Actually, I have around four indicators for each of the three latent variables, but the model 'best' fits only when using one indicator for each.

Regards,
Basel

[baselabuali]

Dear Dr. Jan

Is it possible to share with you more details on my model and data?

Basel

[jmbecker]

We cannot provide individual counselling on how to build an appropriate model. We can only answer general methodological questions that might also be interesting for other researchers.

If you need more input on your specific model, you may either purchase the enterprise license which comes with 8 hours of individual counselling (https://www.smartpls.com/pricing) or attend one of our workshops (https://www.smartpls.com/courses).

[baselabuali]

Thank you Dr. Jan, very clear.

Basel