I am currently working on earthquake risk perception. I have information of perceived probability of occurrence of an earthquake, perceived damage to property, perceived damage to life, perceived level of fear of earthquake (on an ordinal scale). All these four risk perception variables, loads into single latent variable.Let's name the factor score value from factor analysis as a variable Overall Risk Perception. I have information on sociodemographic factors, no of earthquake experienced, time gap from last earthquake experienced etc which I can hypothesize that influence these four risk perception variables. I found from Multiple logistic regression that risk perception parameters have significant relation with four risk perception parameters and overall risk perception.

I would like to hypothesis perceived damage to life and perceived damage to property are related with perceived fear (it would be a two way arrow) and develop a SEM. I do not have any other latent variable. As I have only one latent variable will it be possible to develop Structural Equation Modelling.