Popular Reading Program Raises Questions

Published October 1, 1999

One of most touted remedial reading program nationwide, Reading Recovery┬« (RR), continues to expand, despite controversy over its effectiveness and reports that schools in some districts–including Columbus, Ohio, the first U.S. district to adopt RR, and the Wake County Public School System in North Carolina–are moving away from the program.

Jean Bussell, executive director of the Reading Recovery Council of North America (RRCNA), predicts nationwide figures for the school year ended June 1999 will show 135,000 children enrolled in this first grade tutoring program, a 10 percent increase over last year.

In Columbus, where RR is headquartered at Ohio State University (OSU), only 24 of the city’s 94 elementary schools will use the program in 1999-2000. By contrast, the program’s school participation had peaked at 60 in 1992-1993.

Promoters of the program, aimed at at-risk readers, claim it reduces the need for other remedial programs. Yet 81 percent of Columbus Public School students who finished RR remained eligible for Title I remedial reading, according to a study by the district’s Department of Program Evaluation.

According to a four-year study by the Wake County Public School System, only two of the seven students served in a year by the average RR tutor could read at third-grade level at the end of third grade. RR students “compared to a control group, were just as likely to be retained, placed in special education or served in Title I a year later,” the study concluded.

For each student deemed a long-term success by the program, RR cost about $9,211 beyond regular instructional expenses.

Reading Recovery® was designed by Marie Clay, a New Zealand child psychologist, and brought to the U.S. by OSU in 1984. Since then, it has spread to 10,612 schools across the nation, with 19,570 teachers and teacher leaders trained in RR, according to the RRCNA.

RR’s stated goal is to bring the bottom 20 percent of first-grade readers in their classroom up to the average classroom level. However, when University of Michigan professor Elfrieda Hiebert studied data for 78,000 RR students, she found the average entry-level percentile score of children completing RR was only 34.5, not 50.0.

Other independent studies, including one by Timothy Shanahan, director of the University of Illinois (Chicago) Center for Literacy, and Rebecca Barr, professor of reading and language at National Louis University in Evanston, Illinois, report that children rated as successes by RR fail to learn at a rate comparable to average children in their classroom. RR students who initially outperform children in other remedial programs lose this advantage by third or fourth grade, say Shanahan and Barr, a finding confirmed by Hiebert’s work as well.

Shanahan and Barr report that the average first grader declared “successful” by RR had reached “book level 10” of RR’s 24 book levels.

Children scoring at this level are not reading “authentic text” consisting of a natural pattern of words and sentences. That’s because RR’s books contain “predictable text”–repetitive sentence patterns, context cues, and pictures–that helps children guess at words. Since authentic text is seldom predictable, mastering RR books may not transfer to real reading, say other researchers.

Another study, by the New Zealand Ministry of Education, found that most children who completed RR remained about a year behind their age level in reading and, most significantly, failed to outperform other poor readers who had not had the benefit of RR. The report blamed this failure on RR’s lack of “systematic instruction in word-level strategies”–teaching children to decode printed words.

RR uses “principles and practices very similar to those of Whole Language which has been shown clearly to be a failed instructional innovation,” says Patrick Groff, emeritus professor of education at San Diego State University, a nationally recognized expert on preventing reading failure. Groff and other experts note that Whole Language teaches beginning readers to use context cues to guess at printed words, instead of using phonics skills to decode them.

“The Whole Language approach to reading simply does not work for children with reading disabilities. A structured, phonics-based approach is more likely to help them,” concluded a 13-year study by the National Institute of Child Health and Human Development.

However, in her training guide, RR inventor Clay discourages a structured approach based on phonics, telling tutors that children’s progress in reading is “seriously threatened” if tutors get too “involved in teaching for detail” of print.

While crediting the program with some positive influence on children’s word recognition skills, Groff regards the RR approach as pedagogically unsound. He also thinks the program does not deserve the esteem it has gained in schools, especially when there are other, less costly and better verified, tutoring programs available.

“Only by ignoring the compelling evidence regarding RR’s instructional weaknesses, as well as its relatively high costs, can school boards vote to take money away from other, more deserving school programs, and divert it into RR,” said Groff.