| The purpose of this study was to develop and validate a science assessment tool termed the Life Sciences Assessment (LSA) in order to assess preschool children's conceptions of basic life sciences. The hypothesis was that the four sub-constructs, each of which can be measured through a series of questions on the LSA, will make a significant contribution to the latent construct, thereby causing the LSA to be a significant indicator of preschool children's conceptions of life sciences. Two research questions pertained to this study "What is the contribution of each sub-construct to the latent construct on the LSA?" and "What is the technical adequacy of the LSA in measuring preschool children's conceptions of basic life sciences?" The assessment tool, which consisted of a number of photographs and related questions, was tested with a convenience sample of 124 preschool children from four accredited child care centers in Cincinnati during the 2012-2013 school year. Confirmatory factor analysis was used to evaluate the hypothesized model. Results revealed a good model fit: chi² (1, N = 124) = 0.929, p = .335; NFI = .992; CFI = 1.000; RMSEA = .000; ratio of the chi² to the degrees of freedom = 0.929. In addition, all standardized regression coefficients demonstrated statistical significance at p .3), with moderate to high standardized regression coefficients ranging from .43 to .98. The reliability and validity assessments used to examine the technical adequacy of the instrument indicated that the LSA is a reliable and valid measure for this sample of preschool children: Cronbach's alpha = .733; construct reliability = .714; content validity addressed in stages as required; construct validity established by CFA demonstrating statistical significance at alpha = .001 (2-tailed) as well as practical significance; and successful minimization of potential threats to internal validity. |
