| Abstract: | In this study, a multiple-choice test entitled the Science Process Assessment was developed to measure the science process skills of students in grade four. Based on the Recommended Science Competency Continuum for Grades K to 6 for Pennsylvania Schools, this instrument measured the skills of (1) observing, (2) classifying, (3) inferring, (4) predicting, (5) measuring, (6) communicating, (7) using space/time relations, (8) defining operationally, (9) formulating hypotheses, (10) experimenting, (11) recognizing variables, (12) interpreting data, and (13) formulating models. To prepare the instrument, classroom teachers and science educators were invited to participate in two science education workshops designed to develop an item bank of test questions applicable to measuring process skill learning. Participants formed “writing teams” and generated 65 test items representing the 13 process skills. After a comprehensive group critique of each item, 61 items were identified for inclusion into the Science Process Assessment item bank. To establish content validity, the item bank was submitted to a select panel of science educators for the purpose of judging item acceptability. This analysis yielded 55 acceptable test items and produced the Science Process Assessment, Pilot 1. Pilot 1 was administered to 184 fourth-grade students. Students were given a copy of the test booklet; teachers read each test aloud to the students. Upon completion of this first administration, data from the item analysis yielded a reliability coefficient of 0.73. Subsequently, 40 test items were identified for the Science Process Assessment, Pilot 2. Using the test-retest method, the Science Process Assessment, Pilot 2 (Test 1 and Test 2) was administered to 113 fourth-grade students. Reliability coefficients of 0.80 and 0.82, respectively, were ascertained. The correlation between Test 1 and Test 2 was 0.77. The results of this study indicate that (1) the Science Process Assessment, Pilot 2, is a valid and reliable instrument applicable to measuring the science process skills of students in grade four, (2) using educational workshops as a means of developing item banks of test questions is viable and productive in the test development process, and (3) involving classroom teachers and science educators in the test development process is educationally efficient and effective. |
