Assessing the Physical Activity Questionnaire for Adolescents (PAQ-A): Specific and General Insights from an Ethiopian Context.

The Physical Activity Questionnaire for Adolescents (PAQ-A) has been used in a variety of forms and in a range of countries. This study involves a detailed examination of the PAQ-A to determine its applicability and effectiveness in an Ethiopian setting. We administered the scale to 110 Ethiopian adolescents on two occasions, 5 weeks apart. Data were inspected for features typical of the participants and analyzed to identify interitem correlations, the scale's factor structure, and a range of descriptive statistics concerning composite scores. Most of the scale's items were satisfactorily interrelated according to lenient criteria, and most items loaded on a single factor in exploratory factor analyses. However, a number of the scale's properties were deficient according to stringent or conventionally accepted psychometric criteria. Close inspection of participants' responses highlighted problems in the way the scale is worded, interpreted by participants, and scored. Although the scale does not capture PA as an homogeneous construct, we argue that this is not a problem and neither is its poor test-retest reliability. We make recommendations concerning presentation and scoring of the PAQ-A that are likely to enhance its validity beyond Ethiopia, and we provide a modified version of the scale.