Significance testing on representativeness was also carried out at the ninety percent confidence level.
Six tests of representativeness were undertaken using 2007 data only. Three tests were carried out each on attainment (GCSE results) and on motivation (unauthorised absence). The reference samples were:
• 141 rebuilt schools
• 331 refurbished and rebuilt schools
• All schools in England
Two-tailed paired-sample t-tests were used to test representativeness. The student attainment sample had 89 schools whereas the motivation sample contained 81 schools. The means of the two samples were:
• Not statistically different from that of fully rebuilt schools, i.e. datasets are representative
• Not statistically different from that of refurbished and fully rebuilt schools, i.e. datasets are representative
• Statistically different from the mean for all schools in England (not representative), i.e. datasets are not representative