Development and Standardization of Biology Achievement Test

The general purpose of this study was to develop and standardize an Achievement test to measure student learning in the biology program at the secondary school. Both classical and IRT models were used to address the research objectives of the study. The preliminary instrument consisted of 150 multiple-choice items that on a sample size of 300 male and female students were performed. The final instrument was two parallel forms of 50 items that were performed on a normative sample of 938 male and female students in Shiraz. The Estimated reliability coefficient for internal consistency with test forms was 0.89, and 0.88, respectively. Based on factor analysis; both forms of the test were an overall factor saturated. Results showed there is no significant difference between the mean scores of boys and girls. So standardized and percentile Norms for all subjects were calculated. Findings from the IRT analysis showed that more than %92 of the items are significantly fitted to Three-Parameter Logistic Model. Test information function was a bell-shaped curve and over a wide ability range from -0.5 to +2.5 provides more information. Also, the maximum information was provided at +1.5 from the ability continuum. Based on this, it can be concluded that the Biology Achievement Test provides a more accurate estimate of the True score of subjects whose ability level is higher than the average ability continuum.