Average vs item response theory scores: an illustration using neighbourhood measures in relation to physical activity in adults with arthritis.

OBJECTIVES: Our study had two main objectives: 1) to determine whether perceived neighbourhood physical features are associated with physical activity levels in adults with arthritis; and 2) to determine whether the conclusions are more precise when item response theory (IRT) scores are used instead of average scores for the perceived neighbourhood physical features scales. METHODS: Information on health outcomes, neighbourhood characteristics, and physical activity levels were collected using a telephone survey of 937 participants with self-reported arthritis. Neighbourhood walkability and aesthetic features and physical activity levels were measured by self-report. Adjusted proportional odds models were constructed separately for each neighbourhood physical features scale. RESULTS: We found that among adults with arthritis, poorer perceived neighbourhood physical features (both walkability and aesthetics) are associated with decreased physical activity level compared to better perceived neighbourhood features. This association was only observed in our adjusted models when IRT scoring was employed with the neighbourhood physical feature scales (walkability scale: odds ratio [OR] 1.20, 95% confidence interval [CI] 1.02, 1.41; aesthetics scale: OR 1.32, 95% CI 1.09, 1.62), not when average scoring was used (walkability scale: OR 1.14, 95% CI 1.00, 1.30; aesthetics scale: OR 1.16, 95% CI 1.00, 1.36). CONCLUSION: In adults with arthritis, those reporting poorer walking and aesthetics features were found to have decreased physical activity levels compared to those reporting better features when IRT scores were used, but not when using average scores. This study may inform public health physical environmental interventions implemented to increase physical activity, especially since arthritis prevalence is expected to be close to 20% of the population in 2020. Based on NIH initiatives, future health research will utilize IRT scores. The differences found in this study may be a precursor for research on how past and future treatment effects may vary between these two types of measurement scores.

The effect of average scores in reading, mathematics and science on innovation and income: A quantitative analysis for a group of countries.

The purpose of this paper is to The goal of this study is to empirically examine how innovation and income are affected by PISA scores-average scores in math, science, and reading. The specific objectives are threefold. First, we develop a conceptual model that describes the connection between income, innovation, and educational quality as determined by PISA scores. Second, we perform an empirical analysis to see if educational quality as assessed by the PISA scores has an impact on innovation.Third, we also test whether the PISA scores have a positive effect on income (GDP per capita). Using panel mode, we find that education quality has a significant and positive effect on innovation and income. Particularly high math test scores have been found to be an important subject that increases per capita income and innovation.