Yes, we can. You are not required to model every variable as latent in an SEM. You can also include manifest (observed) variables such as age, gender, and income into the model (i.e., variables for which you do not have multiple indicators and/or for which you can safely assume that they are free or nearly free of measurement error).
Sanjay Krishnamurthy Yes, in a structural equation model, continuous demographic data (such as age, income, and job experience) can be used with latent variables (SEM). Latent variables are variables that are not directly observed but are inferred from the correlations between other variables that are observed or measured.
Continuous demographic variables can be employed as either observable variables or latent variables in SEM, depending on the research topic and analytic aims. For example, if the purpose is to investigate the link between age and another end variable, such as work satisfaction, age may be utilized as an observed variable. Alternately, if the purpose is to investigate the underlying construct or component that is reflected, age might be employed as a latent variable.
In SEM, it is critical to carefully analyze the variables and their measurement, as the findings of the analysis might be sensitive to the variables' measurement and the assumptions made about their correlations.