Based on the need for, suggestions about the construction, and existing measures serving as models found in the literature both inside and outside of science education, the “Children's Science Curiosity Scale” has undergone six versions using four different samples of fifth graders. Respectable internal consistency (alpha) and test‐retest reliabilities have been calculated. Concurrent validity has been demonstrated by significant positive correlations with another recognized curiosity measure, and by way of significant differences between students who were interested in science and students uncertain about interest in science. Somewhat weak predictive validity has been decided by way of significant positive correlations with students' semester science grades. Construct validity has been described and established by eight judges using six criteria, and supported through the use of factor analysis where four underlying factors were hypothesized as characteristics of science curiosity. Sex differences were also explored where significant differences were not found between the genders. Suggestions have been made concerning future attempts at instrument refinement, establishing conceptual validity, future research involving other variables, and classroom use in a variety of contexts.

The purpose of this investigation has been to explore whether differences existed between gifted and nongifted fifth graders and between genders and related subgroups with respect to attitudes toward science. Both groups (N = 25) were matched on the demographic characteristics of school‐site, race, sex, and socio‐economic background. Gifted students were found to have more positive attitudes toward science than nongifted students; however, no significant differences were found. In all cases, boys (all boys, gifted boys, and nongifted boys) exhibited more positive attitudes toward science; again, no significant differences were uncovered between the boys and their counterpart group or subgroups. The item which consistently reflected the most positive rating (gifted students, all boys and gifted boys, and all girls and nongifted girls) was “usefulness of things done in science class.” Items where discrepancies surfaced included “usefulness of science when playing at home” where nongifted students and gifted girls were significantly more positive than their counterparts, and “spending more time doing science experiments” where all boys and gifted boys were significantly more positive than their counterparts.

The intent of this study was to investigate possible relationships among the constructs of attitudes toward science, interest in science, science curiosity, and self‐concept of science ability. Data were collected from 228 sixth‐grade students by way of four Likert‐type instruments, each of which measured a given attribute. Statistical procedures included univariate and multivariate analysis of variance, product‐moment correlations, canonical correlations, multiple regression analysis, and factor analysis. The results indicated that attitudes toward science, interest in science, and science curiosity are highly related. Implications for classroom science teaching and suggestions for future research have also been discussed.

No abstract

We do thank Professor Gardner for his interesting and thought provoking comments. We shall react to the four points that he has raised.