12. Summary

The results from the Pilot Study provide support both for the content validity and the construct validity of the Numeracy scale. The 40 items selected for the Main Study cover the Numeracy domain as envisioned. They reflect the facets contained in the definition of measurable numerate behavior, include items at all levels of difficulty, and satisfy psychometric requirements common in international comparative studies.

In addition, analysis of the pilot data suggest that "non-cognitive" items, including several item clusters developed by the Numeracy team, can serve as useful covariates of observed performance on Numeracy items and of numeracy skill levels. Some of these items ask about general attitudes concerning mathematical tasks and about recollections of mathematics learned in school settings. Other items ask about a respondent's involvement with particular numeracy tasks on the job, about self perceptions as to whether the respondent's math skills are sufficient to accomplish what is required at work, and examine certain beliefs, attitudes, and practices related to numeracy. Another item asks for a self-assessment of the extent to which a respondent used a calculator during the Numeracy block.

Overall, the forty items chosen for Numeracy assessment can be used to describe the distribution of numeracy skills in the population as well as in important sub-groups. In addition, the supplemental non-cognitive items included in the Background Questionnaire can shed light on possible antecedents and consequences of the numeracy skills that are of interest to policy makers and other audiences.