Introducing equating methodologies to compare test scores from two different self-regulation scales.

Area of Science:

• Health behavior research
• Health education research
• Psychometrics

Background:

• Standardizing measurement tools across studies is crucial for comparing intervention effectiveness but is difficult in practice.
• Researchers often prefer tailored measurement tools specific to their interventions.
• Comparing results from studies using different scales for the same construct is a significant challenge.

Purpose of the Study:

• To demonstrate the utility of item response modeling linking methodology for comparing different measurement scales of the same underlying construct.
• To empirically apply this methodology using the Treatment Self-Regulation Questionnaire (TSRQ) in the context of diet improvement.
• To assess the impact of common items on the comparability of different scales.

Main Methods:

• Item response modeling linking methodology was employed.
• Two different eight-item Treatment Self-Regulation Questionnaire (TSRQ) scales measuring diet improvement were compared.
• The number of common items between the scales was varied to assess its effect.

Main Results:

• Two eight-item TSRQ scales can be reliably linked if they share at least four items.
• The number of linking items did not impact the reliability of the results.
• Varying the number of linking items significantly affected the relative rating when compared to a 15-item scale.

Conclusions:

• Item response modeling linking methodology is a valuable tool for comparing results across studies using different versions of a scale to measure the same construct.
• This method enhances the ability to conduct meta-analyses and synthesize findings in health behavior and health education research.
• The number of common items is a critical factor in determining the comparability and relative scaling of different measurement instruments.