The American Institute for Research (AIR), one of the country’s leading research organizations, looked at nine years of historical data in a total of 229 schools in NYC and Georgia, matching TC schools demographically with non-TC schools, and found that the TC partner schools made more improvements in student achievement than did the matched schools that did not partner with TCRWP. TCRWP schools made statistically-significant improvements starting in their second year and those improvements grew steadily larger over time.
Main Results
Changes in reading achievement of students in TCRWP classrooms compared with that of students in comparison classrooms are represented graphically, below. Figure 1 shows the pre- and post-trends of test scores for treatment schools and for matched comparison schools before and after TCRWP was adopted by treatment schools. The vertical line at 0 indicates the last test administration before treatment schools started implementing TCRWP. As the figure shows, there are no apparent differences in test scores between treatment and matched comparison schools in the 3 years prior to TCRWP implementation (i.e., years -2, -1, 0). Relatedly, the scores for both groups display the same trend throughout the 3 years, in that the trends are both relatively constant. However, beginning in the first TCRWP implementation year, we observe that TCRWP treatment schools’ test scores increased, while comparison schools’ scores remained relatively constant. In later years, the two groups’ scores follow relatively similar trends, but in all post-implementation years, TCRWP treatment schools’ scores remain higher than those scores of the comparison group in post-implementation years, and, in general, the differences widen over time. By the 5th and 6th years after the implementation of TCRWP, treatment schools’ scores have become higher than those scores of the comparison schools by 0.21 standard deviations.
Figure 1. Pre- and Post-Trends in Standardized ELA Test Scores for Treatment and Matched Comparison Schools, Descriptive Data
Note. Data points represent standardized average test scores from the 51 treatment schools (blue) and the 178 matched comparison schools (red). All schools in the sample have test score data for at least 1 year after implementation. However, the results for later years are based on fewer schools (e.g., seven schools that began implementing TCRWP in 2012–13 have seven years of post-implementation data).