The Failure of Reading First

Stephen Krashen
Reading Today (August/September, 2008)

The failure of Reading First

According to the Reading First Impact Study: Interim Report, released April 30, 2008, children in Reading First did no better than children in comparison groups. In the media descriptions, as well as in the report itself, there have been several serious omissions.

Not the first failure

This isn't the first failure of Reading First. The current dismal results are consistent with all previous analyses. Members of the administration have repeatedly claimed that National Assessment of Educational Progress (NAEP) reading scores for fourth graders are at an all-time high, but a brief look at the data shows that nearly all of the improvement took place before Reading First was implemented.

Higher scores at grade levels on state tests since the implementation of Reading First have been interpreted as evidence for its effectiveness, but improvement was present before Reading First, and a number of studies have shown that Reading First did not increase the rate of improvement. Also, a recent analysis has shown that fewer children were classified as proficient at grade 3 than at grade 1. The decline appears to be greater for more recent cohorts of children on tests of comprehension, but not on tests of "fluency" (i.e. those starting in 2005 compared to those starting in 2004). If later cohorts represents a purer application of the principles of Reading First, with its focus on intensive, systematic phonics, the results are consistent with previous research showing that the effect of intensive phonics is obvious only on tests in which children pronounce words, and is much weaker on tests in which children have to understand what they read.

Also, American fourth graders did not improve between 2001 and 2006 on the Progress in International Reading Literacy Study (PIRLS) reading test. In addition, there has been no reduction of the gap between children from high- and low-income families on any measure. Reading First has failed every time it has been put to the empirical test.

Less effective than comparison methods

The Reading First Impact Study (and previous analyses) actually showed that Reading First was less effective than the comparison methods. Reading First students had an additional 10 minutes per day of instruction on the elements of reading that Reading First assumed were crucial (derived from the National Reading Panel report). That's the equivalent of an extra six weeks of reading instruction every year (assuming an hour a day spent on reading in the regular school day).

Even if Reading First were only mildly effective, the extra time should result in higher reading scores. It didn't. This means that time was taken from other subjects and activities, as well as recess, and students got nothing in return. This must be one of the great failures of educational research.