Skip Navigation
Effectiveness of Reading and Mathematics Software Products: Findings from the First Student Cohort
NCEE 2007-4005
March 2007

Collecting Achievement and Implementation Data

The study's analyses rely mostly on data from student test scores, classroom observations, and teacher questionnaires and interviews. The study also collected student data items from school district records and incorporated data about districts and schools from the National Center for Education Statistics' Common Core of Data.

To measure effects, the team administered a student test in the fall and spring of the 2004–2005 school year. The team used the Stanford Achievement Test (version 9) reading battery for first graders, the Stanford Achievement Test (SAT-10) reading battery for fourth graders, and the SAT-10 math battery for sixth graders. These tests were administered in fall 2004 and spring 2005. The team also used the Test of Word Reading Efficiency (TOWRE), a short and reliable one-on-one test of reading ability, for first graders to augment measures of reading skills provided by the SAT-9 (Torgesen et al. 1999).

To measure algebra achievement, the study selected Educational Testing Services' (ETS) End-of-Course Algebra Assessment (1997). Because baseline measures of algebra knowledge were not available or were considered unsatisfactory, the study worked with ETS to separate its assessment, which essentially is a final exam, into two components that had equal levels of difficulty. The study randomly selected classrooms either to take part A in the fall and part B in the spring or to take B in the fall and A in the spring. Splitting the test in this way meant that the full test was administered in both the fall and the spring, but each student took only half of the test at each point. The team also collected scores on district achievement tests if these data were available. The study's administration of its own test provided a consistent measure of achievement across varied districts and schools, but examining findings based on district tests provided a useful check on the robustness of the findings.

Classroom observations were the study's primary basis for assessing product implementation. An observation protocol was developed in spring 2004, and videotapes of classrooms using products were gathered and later used for observer training. The observation protocol was designed to gather similar information in both treatment and control classrooms and across the different grade levels and subject areas in the study. In addition, the protocol was designed to focus on elements of instruction and implementation that could be observed reliably. Each classroom was visited three times during the school year, and observers used the protocol for each observation, which lasted about 1 hour. Observations were complemented by a teacher interview that gathered information about implementation issues. Background information about teachers was also gathered from a questionnaire that teachers completed in November and December 2004.