Assessment of handedness using latent class factor analysis

Area of Science:

• Psychology
• Human Movement Science
• Statistical Modeling

Background:

• Handedness is often studied as a latent construct, utilizing various statistical models like latent class, factor, and mixed Rasch models.
• Previous research has employed qualitative and continuous latent variables to model handedness, but a specific approach integrating ordered categories requires further exploration.

Purpose of the Study:

• To investigate the utility and efficacy of treating handedness as a qualitatively scaled latent variable with ordered categories using latent class factor models.
• To compare the performance of the latent class factor model against traditional latent class analysis for handedness assessment.

Main Methods:

• An exploratory analysis was conducted on a dataset comprising 2236 young Italian sportspeople and their hand preference across 10 distinct tasks.
• Latent class factor models were applied to analyze handedness patterns, with latent class analysis performed for comparative purposes.
• A cross-validation procedure was implemented to assess the robustness and generalizability of the model findings.

Main Results:

• The latent class factor model demonstrated the best fit to the observed handedness patterns compared to other models.
• Individuals were successfully assigned to one of four distinct, ordered levels of handedness.
• A quantitative index of left-handedness was computed for each individual, considering the task-specific effects on hand preference.

Conclusions:

• The latent class factor approach provides a robust and effective method for assessing handedness.
• This model allows for nuanced categorization of individuals into ordered handedness levels and yields a quantitative measure of left-handedness.
• The findings support the adoption of the latent class factor model for comprehensive handedness assessment in research and practice.