From: Stephen Krashen <skrashen@yahoo.com> The US Department of Education is claiming that Reading First is working. This is on the basis of a recent report claiming that the percentage of third graders meeting or exceeding the proficient level on tests of reading comprehension increased by 12% between 2004 and 2006. I have examined this data (available on the Department of Education website). For the 30 states with test scores available, I calculated an average increase of 6.7% in the percentage of third graders scoring at the proficient level or higher between 2004 and 2006. (If the number of children in each state is considered, the "weighted mean" is 6%, half the figure reported by the Department of Education.) Four states showed huge gains, about 20% or more (Indiana, 26%, New Hampshire, 20.5%, South Carolina, 21%, Virginia, 19.5%), but these four included only 4.5% of the total number of students who took the test. Three of the states with the most students tested did not do as well: California gained 4%, Texas 2% and Penn decreased nearly 10%. Combined, these three states make up about half of the total number of children tested. When we consider that children in Reading First get 100 extra minutes of reading instruction per week, and Reading First teachers get significantly more professional development, these are not "impressive gains," as the Department of Education Press Release of April 19 claims. Also, as others have pointed out, no comparison group was included in the analysis. The Department of Education refuses to consider any research that does not meet strict "scientific" standards, yet has no problem concluding from this data that Reading First "is working." Stephen Krashen Note: I limited my analysis to grade three, as these children have had the greatest chance to profit from Reading First. I only considered reading comprehension, as previous studies have shown that heavy phonics-based methods show an advantage on tests of word reading, but not on tests of reading comprehension given after grade one (Garan's analysis of the National Reading Panel report).