The specific focus of this research was to determine how the outcomes on two alternative forms of assessment (multiple-choice and hands-on/manipulative) for science process skills were related when students were grouped on the basis of sex, race/ethnicity, and poverty level. Subjects were 1,381 fourth graders in a culturally diverse city school district in New York state. Also explored was the relationship between the reading scores of students and their scores on the two alternative science assessments and the effect of the presence of an elementary science specialist on the student scores. The reliability of New York's Elementary Science Program Evaluation Test (ESPET) was also considered. All students performed better on the hands-on test, and the traditional gap associated with socioeconomic or racial/ethnic status was reduced by the hands-on test. Economic status appeared to have an effect on science achievement, and reading had a greater effect than poverty on science achievement. Results supported the reliability of the ESPET, but did not support the importance of an elementary science specialist to student achievement. Results show that all students can learn science given appropriate instruction. Thirteen tables and 17 figures present study findings. (Contains 48 references.) (SLD)