Latent versus Observed variables

Consider the following questions asked by a researcher in a questionnaire on Social Acceptance:

  1. The product should be approved by my friends.
  2. The product should be popular in my circle of friends.
  3. The product should make a good impression on people who are important to me.
  4. The product should be liked by people who are important to me.
  5. The product should be liked by people who are important to me.

Source: https://doi.org/10.1016/j.dib.2017.04.054

I am pretty sure you have come across several of these examples in any survey questionnaire. In this example, Social Acceptance is measure by five items.

Why?

This is because Social Acceptance by itself is too broad and the researcher cannot measure it directly, so he or she is tasked to design a set of questions/items (that have been empirically tested and conforms to theory) to measure Social Acceptance. These sets of questions/items are called observed/manifest variables. In Structural Equation Modeling (SEM), they are usually called indicators. Social Acceptance, on the other hand, is known as a construct/factor/latent/unobserved variable. Other latent variables include Quality, Hedonic Motivation, Pleasure, Attitude. and so on.

Did you find this page helpful? Consider sharing it 🙌

Next
Previous