The model of structural relationships is what one really wants to estimate. It contains the effects and relations between the constructs, which would normally be latent variables. Latent variables are

The model of structural relationships is what one really wants to estimate. It contains the effects and relations between the constructs, which would normally be latent variables. Latent variables are characteristics that one would like to measure; however, it cannot be readily observed and it is free from measurement error (Bollen, 2002). Latent variables are measured by observable indicators called manifest variables. These latent variables can be exogenous or endogenous. Latent variables are non-observed random entities, which do not have a specific unit of measurement and, therefore, any estimate is subject to an associated error (Kuttner, 1994). An example of how I might go about developing a latent variable that may be of interest to me would be in such a way as follows. Cultural competence in health professionals is a latent variable. In case of having a valid instrument to measure this variable, the assignment of any value would have, unquestionably, an associated margin of error and its unit of measurement could vary between studies since there is no consensus in its comparison parameter (Bollen, 2002). Therefore, the question arises of how to properly access latent variables in health research. The answer lies in the use of a group of variables or indicators that can be obtained directly and that, in turn, are theoretically associated with the latent variable of interest, such as, for example, questions in a questionnaire or items in a questionnaire or poll (Kuttner, 1994). Thus, it can be modeled the probability of answering a set of questions or indicators observed given a certain level in the latent trait.

References

Bollen, K. A. (2002). Measuring Latent variables in. Annual Review of Statistics, 53(1), 605-634.

Kuttner, K. N. (1994). Estimating potential latent variable. Journal of business & economic statistics, 12(3), 361-368.

It is true that we compute latent variables by taking a linear combination of multiple observed, or measured, variables. In your response for example you discuss interest in cultural competence as a latent variable. What are some different measured variables you believe would help you to create a good latent variable that represents cultural competence?

• Cite all sources in APA format.
• Attach a Turnitin.com Report.