Latent variable with one indicator: is it possible?

[marianamr]

Hi!

I'm testing a model and I have one predictor (age) that I want to include in the model. Is it possible to include one latent variable with just one indicator like this?

Thanks

[PhilippGG]

My guess is no, it is not. Latent Variables are usually complex theoretical constructions, which are hard to measure. With only one variable you would probably explain only one aspect of the latent variable, so your measure would lack validity. Furthermore I read that in PLS, the more indicators you have, the better is your forecast.

so long, Philipp

[casander]

Hi,

I dont understand why you would use a latent variable "age"? If it is measured in years (which I assume), there is no need for a measurement model with multiple indicators since it is not abstract but rather concrete with everybody understanding what you mean.

You can include "age" as a single indicator for the "latent" variable "age" in PLS, which in turn is no longer latent... The weight/loading in that case equals 1, so it does not matter if reflective or formative.

maybe this reference is useful:
C. Fuchs / A. Diamantopoulos (2009): Using single-item measures for construct measurement in management research, in: Die Betriebswirtschaft, Vol. 69, No. 2, p. 195-210. (english article in german journal)

cheers

[ghozali]

Hi,
If you want to use latent with one indicator, just make one indicator in the latent variable but the indicator age should be in formative form