TASC Impact on Student Learning and Classroom Practices
3. Design, Data & Analysis
For the preliminary analyses to be presented, student 2006-07 End-of-Grade (EOG) state achievement scores in mathematics and reading were obtained from three partnering school systems for grades 3 through 8 along with student demographic data and assigned science teacher. Student 2006-07 EOG state achievement scale scores in science were limited to grades 5 and 8 since students were tested in science only in these grades. Student responses to the 2006-07 EOG survey questions regarding classroom science teaching practices in grades 3 through 8 were also used. The professional development history of each of the teachers in the targeted school systems was collected from school system human resource files dating back to 1998. This professional development data for the teachers was categorized as either TASC-provided training, other science training (not TASC-provided) or non-science training. A teacher who received professional development by a TASC staff member in the use of a science kit or in inquiry-based instruction was placed in the TASC-provided training category. A teacher who received professional development in any science topic or in the use of a science kit but not by a TASC staff member was placed in the other science training category. A teacher not documented to have received any science training was placed in the non-science training category. Descriptive statistics were compiled for each of the variables of interest, and multiple regression analyses were conducted to assess the hypotheses of interest. To compile the project data, several school system databases were combined targeting fields of interest. For instance, student data needed to be matched correctly to the science teacher and human resource data coded for teacher training categories. The dataset used in these analyses includes the responses of 10,984 students merged with one and only one science teacher id to whom they were assigned. The project data are currently in the process of validation. Future analyses will include hierarchical linear regression models to account for nested variables and will likely include student achievement and science survey data from 2002-2003 through 2007-2008.
