A Simulation-Based Comparison Between Estimated Latent Scores and Sum Scores

• In Event: AERA Poster Session Saturday 3:45 pm
  In Invited Poster Session: Division D Graduate Student In-Progress Research Poster Session

Abstract

Composite scores are common in research, yet the choice between sum scores, factor scores, imputed scores, and structural equation modeling remains controversial. Extending Liu and Pek (2024) work, this simulation evaluates these methodologies’ performance to recover parameters in a mediation model. The design manipulates measurement error, scale length, model misspecification, and sample size. Preliminary results based on bias and MSE indicate that while scale length has negligible impact, the efficacy of specific scoring methods depends significantly on the error form and the scoring method. These findings provide empirical guidelines for balancing computational simplicity with estimation accuracy in latent variable modeling.

Authors

• Tonatiuh Xochihua Tlecuitl, University of California - Los Angeles

• Craig K. Enders, University of California - Los Angeles

• Yi Feng, University of California - Los Angeles