This is one in a series of 15 two-page, evidence-based framing memos on pressing education and social issues prepared by MDRC for the incoming Obama Administration and the new Congress.

Bottom Line

The Reading First program, a central part of No Child Left Behind, has been the largest recent federal effort to improve elementary school reading instruction through support for effective curricula, diagnosis and assistance for struggling readers, and training for teachers. Although funding for Reading First has been cut dramatically and the future of the program remains in doubt, policymakers still face key choices about what to do to improve early reading. Recent research finds that federal support for teacher professional development can improve two potentially critical precursors to improved reading by students: (1) teacher knowledge of scientifically-based reading instruction and (2) teacher instruction and practice in the classroom. But it also suggests that impacts on student reading achievement are more difficult to realize.

What Do We Know?

Much recent federal support — within Reading First and other special initiatives — has focused on training for the existing teachers in low-performing schools that serve significant numbers of low-income students. Much research has shown that the preparation, knowledge base, and experience of teachers in low-performing schools is on average weaker than in more successful schools and/or in schools serving higher-income communities. Though there are many initiatives — funded by government and foundations — to improve “preservice training” in universities and to change the allocation of teachers across schools, the existing teachers in low-performing schools remain the central input into children’s education, meaning that professional development for them is critical.

Reading First and other recent large-scale federal professional development (PD) initiatives have emphasized scientifically-based instructional practices found to produce improvements in student reading achievement. These practices, based on the findings of the National Reading Panel, emphasize instruction in five key components of reading instruction: phonemic awareness (sound recognition in words), phonics (letter-sound correspondence), vocabulary, fluency, and comprehension.

Recent federal initiatives have produced improvements in teacher knowledge of how children learn and in teachers’ use of desired instructional practices. In the two recent evaluations, one of Reading First and the other of two reading PD interventions, federal support for PD did change teacher instructional practices. For example, both studies found it led to greater emphasis on explicit instruction in the key components of reading instruction. And in the reading PD study, the enhanced PD led to increases in teacher knowledge of scientifically-based reading instruction methods (the Reading First study did not measure teacher knowledge).

But so far there is limited evidence of PD efforts creating overall improvements in student reading achievement. However, both the Reading First study and the reading PD study measured reading achievement rather narrowly, using standardized tests that emphasized — or exclusively focused on — reading comprehension rather than precursor skills, such as phonics and other “word level” skills that were also a part of the PD tested. However, the use of a comprehension-oriented outcome measure in the evaluations was consistent with the emphasis of many state and local accountability systems in education.

There are suggestions of a lesson about better targeting of PD services: Reading First produced positive impacts on student reading achievement in the second cohort of grantees. These schools initially were using fewer of the intended instructional practices, had lower levels of achievement before the intervention, and coincidentally also received more federal support from Reading First. In this group of schools, impacts on instruction were greater and impacts on student achievement also appeared. While not definite, this supports a hypothesis that targeting schools that aren’t as far along in adopting the desired instructional approach could produce impacts on student achievement.

Impacts on intermediate steps, such as teacher knowledge and instructional practice, may not have been large enough to affect student reading achievement. In addition to the targeting hypothesis, exploratory analysis in the reading PD study found that the changes in teacher knowledge produced by the PD were smaller than prior research suggests is needed to boost student achievement.

Perhaps one-year interventions are not enough. The impacts on teacher knowledge and teacher practice of the one-year PD effort in the reading PD study disappeared in the second year. However, in Reading First, PD was provided for multiple years, and the study found impacts in each year it examined. Continuing reinforcement and support may be particularly important in low-performing schools, where teacher turnover is often 20 percent or more a year.

But greater intensity of PD alone is not the answer. The reading PD study tested unusually intense forms of PD: a multi-day summer institute and follow-up seminars versus PD with those features plus in-school coaching throughout the school year. The differences in impacts on teacher instruction were not large enough to be statistically significant and neither had impacts on student achievement.

Early results on a comprehensive “induction” program for new teachers are disappointing. A recent study by Mathematica Policy Research found that such a program can be successfully implemented to increase mentoring of new teachers. However, the program had no first-year impacts on teacher reading instruction, student reading test scores, or teacher mobility/retention.

Where Do We Go from Here?

Even if these PD efforts have not produced overall impacts on student reading achievement, they have been able to improve teacher knowledge and instruction. Federal support through Reading First and in special PD initiatives has demonstrated that it is possible to fairly quickly change teacher instruction and teacher knowledge, suggesting that further efforts to improve PD could be worthwhile. However, improving student achievement may require a longer timeframe or other changes.

Exploratory analyses suggest that focusing on multi-year PD programs and strategies that target needier schools may show promise.

Continue to experiment with innovative ways to deliver PD and related support in large school districts. For example, recent findings illustrate the need for new ways to deliver teacher coaching that will increase its power to improve instruction.

Key References

The Reading First Impact Study Project

Professional Development in Reading Study Project

Glazerman, S., and colleagues. 2009. Impacts of Comprehensive Teacher Induction: Results From the First Year of a Randomized Controlled Study. Report submitted to the U.S. Department of Education. Princeton, NJ: Mathematica Policy Research, Inc.

National Reading Panel. 2000. Teaching Children to Read: An Evidence-Based Assessment of the Scientific Research Literature on Reading and Its Implications for Reading Instruction. Washington, DC: Author.