The calibration of a sit-ups task using the Rasch Poisson Counts model

Area of Science:

• Sports Science
• Psychometrics
• Biostatistics

Background:

• Standardized physical fitness tests are crucial for evaluating youth physical activity.
• The Rasch Poisson Counts model offers a robust framework for item calibration and person ability estimation.
• Previous research has highlighted the need for accurate calibration of fitness assessments in diverse populations.

Purpose of the Study:

• To calibrate a nationally utilized sit-ups test using the Rasch Poisson Counts model.
• To assess the model-data fit of the Rasch Poisson Counts model for the sit-ups test.
• To compare performance metrics between genders and age groups on the calibrated sit-ups test.

Main Methods:

• Utilized the Rasch Poisson Counts model for test calibration.
• Analyzed data from 8,723 participants (4,486 girls, 4,237 boys) aged 10-18.
• Evaluated model-data fit across different ability levels and demographic groups.

Main Results:

• The sit-ups task difficulty was estimated at -2.80, suitable for most participants (ability range .09-1.39).
• Boys and girls, as well as different age groups, were compared using a unified metric.
• Model-data fit was superior for higher ability levels compared to lower ability levels.

Conclusions:

• The Rasch Poisson Counts model provides a viable method for calibrating the sit-ups test.
• Discrepancies in model-data fit at lower ability levels may stem from violations of model assumptions regarding performance speed.
• Further investigation is needed to refine the model for participants with lower physical fitness levels.