Using Data From the Idaho Achievement Test as a Tool for School Improvement

Publication Date


Type of Culminating Activity


Degree Title

Doctor of Education in Curriculum and Instruction


Curriculum, Instruction, and Foundational Studies

Major Advisor

William Parrett, Ph.D.


Phil Kelly, Ph.D.


Scott Willison, Ph.D.


Roger Stewart, Ph.D.


The purpose of this three stage study was to investigate how elementary teachers and principals are using the data from Northwest Evaluation Association’s Idaho Standards Achievement Test (Idaho) and the Measures of Academic Progress (California) to improve curriculum and instruction for students. Stage I was conducted with nine principals and forty-five elementary teachers. Teachers participated in focus group interviews and principals were interviewed one on one. Results from Stage I address key differences between teachers in high and low growth grade levels in the knowledge and leadership skills of the principal, the role of the building specialists, teaming, a schedule that focuses on academics, flexible grouping, and curriculum at the appropriate skill level.

Stage II of the study was the development and pilot of a thirty question instrument based on the findings from Stage I. The purpose of this instrument was to predict teacher membership in a high or low growth group based on growth in student achievement in reading between the fall and spring test administrations. The web-based questionnaire was piloted with the focus group participants to observe if the questionnaire would predict with a known sample. Differences in raw percentages between high and low growth teachers were reported as well as chi square results.

Stage III, an instrument validation study, was conducted with six hundred thirty certified elementary classroom teachers and specialists. An administration of the web-based questionnaire with a larger sample in a different school district was conducted to study whether or not the questionnaire was a good predictor with a different sample. A chi square analysis was run to initially select a set of predictor variables. This was followed by fitting a hierarchical linear model to the data which accounted for dependence across teachers in the same school (Stokes et al., 2000; Hardin & Hilbe, 2003).

Files over 30MB may be slow to open. For best results, right-click and select "save as..."