Hierarchical Models of Attitude

Area of Science:

• Social Sciences
• Psychology
• Philosophy of Science

Background:

• Distinguishing levels of abstraction is crucial in philosophy of science.
• Social sciences are increasingly adopting hierarchical models for complex phenomena.
• Previous estimation limitations hindered the use of hierarchical models.

Purpose of the Study:

• To illustrate the application and estimation of hierarchical models in social sciences.
• To demonstrate the utility of hierarchical models using the structure of attitude as an example.
• To compare hierarchical models with non-hierarchical approaches for conceptualizing and estimating collinear constructs.

Main Methods:

• Utilized advancements in statistical estimation techniques (Bentler & Weeks, 1979; Joreskog & Sorbom, 1981).
• Applied hierarchical modeling to the structure of attitude.
• Compared hierarchical models against non-hierarchical models.

Main Results:

• Hierarchical models of attitude demonstrated superior model fit compared to non-hierarchical models.
• The hierarchical approach yielded more interpretable results for attitude structure.
• Hierarchical models provide a robust method for conceptualizing and estimating models with collinear constructs.

Conclusions:

• Hierarchical models represent a significant advancement for social science research, offering improved analytical capabilities.
• The application to attitude structure highlights the practical benefits of these models.
• Researchers should exercise caution regarding the unique determination of higher-order models.