Development of a science achievement evaluation instrument for a school garden program

School gardens show promise as a tool for developing science process skills through real-world investigations. However, little research data exist attesting to their actual effectiveness in enhancing students' science achievement. The purpose of this study was to develop three cognitive test instruments for assessing science achievement gain of third, fourth, and fifth grade students using a garden curriculum. The development of the test instruments occurred in three phases: 1) an initial set of test instruments which served as a prototype for length, scope, and format; 2) an adapted set of test instruments which were piloted; and 3) a final set of test instruments which were used for the assessment of the school gardening curriculum. The final Cronbach's alpha reliability for the final set of test questions was 0.82, indicating an acceptable level of internal consistency. Content validity of the test instruments developed for this study was established based on the science content standards specified in the Texas Essential Knowledge and Skills (TEKS) for each grade level along with the gardening curriculum, as well as the Science Scope and Sequence documents for Temple, Texas Independent School District (ISD). Construct validity was established for the testing instruments by soliciting help from various curriculum experts from the Temple ISD.