Average Impacts on Classroom Reading Instruction, Key Components of Scientifically Based Reading Instruction, and Student Reading Achievement
Exhibit ES.3 reports average impacts on classroom reading instruction and student reading
comprehension pooled across school years 2004-05 and 2005-06 and 2006-07.6 Exhibit ES.4 reports
average impacts on key components of scientifically based reading instruction from spring 2007. Exhibit ES.5 reports the average impact on first graders' decoding skills from spring 2007. Impacts were
estimated for each study site and averaged across sites in proportion to their number of Reading First
schools in the sample. Average impacts thus represent the typical study school. On average:
- Reading First had a statistically significant impact on the total time that teachers spent on the five essential components of reading instruction promoted by the program in grades one and two.
- Reading First had a statistically significant impact on the use of highly explicit instruction in grades one and two and on the amount of high quality student practice in grade two. Its estimated impact on high quality student practice for grade one was not statistically significant.
- Reading First had no statistically significant impacts on student engagement with print.
- Reading First had a statistically significant impact on the amount of professional development in reading teachers reported receiving; teachers in RF schools reported receiving 25.8 hours of professional development compared to what would have been expected without Reading First (13.7 hours). The program also had a statistically significant impact on teachers' self-reported receipt of professional development in the five essential components of reading instruction; teachers in RF schools reported receiving professional development on an average of 4.3 of 5 components, compared to what would have been expected without Reading First (3.7 components).
- A statistically significantly greater proportion (20 percent) of teachers in RF schools reported
receiving coaching from a reading coach than would be expected without Reading First. The
program also had a statistically significant impact on the amount of time reading coaches
reported spending in their role as the school's reading coach; coaches in RF schools reported
spending 91.1 percent of their time in this role, 33.5 percentage points more than would be
expected without Reading First (57.6 percent).
- Reading First had a statistically significant impact on the amount of time teachers reported
spending on reading instruction per day. Teachers in RF schools reported an average of 105.7
minutes per day, 18.5 minutes more than the 87.2 minutes that would be expected without
- Reading First had a statistically significant impact on teachers' provision of extra classroom
practice in the essential components of reading instruction in the past month; the impact was
- There were no statistically significant impacts of Reading First on the availability of
differentiated instructional materials for struggling readers or on teachers' reported use of
assessments to inform classroom practice for grouping, diagnostic, and progress monitoring
- Reading First had no statistically significant impact on students' reading comprehension
scaled scores or the percentages of students whose reading comprehension scores were at or
above grade level in grades one, two or three. The average first, second, and third grade
student in Reading First schools was reading at the 44th, 39th, and 39th percentile respectively
on the end-of-the-year assessment (on average over the three years of data collection).
- Reading First had a positive and statistically significant impact on average scores on the
TOSWRF, a measure of decoding skill, equivalent to 2.5 standard score points, or an effect
size of 0.17 standard deviations (See Exhibit ES.5). Because the test of students' decoding
skills was only administered in a single grade and a single year, it is not possible to provide
an estimate of Reading First's overall impact on decoding skills across multiple grades and
across all three years of data collection, as was done for reading comprehension.
6 Except for student engagement with print (STEP), which is pooled across the 2005-06 and 2006-07 school years only.