ABSTRACT
We noted in chapter 1 that the original plans for our large studies called for samples of 216 students, 54 in each of Grades K-3, stratified according to SES, achievement level, and gender. Our first two studies (on shelter and clothing) did in fact reflect these plans. However, analyses of their data led us to conclude that it would not be cost effective for us to continue to sample systematically from communities representing three different SES levels. In this chapter, we explain that decision in more detail and also present our findings relating to student achievement level and gender.
