The Torgesen Study

“Sixty children with severe reading disabilities… received 67.5 hours of one-to-one instruction in two 50-minute sessions per day for 8 weeks. …  Within 1 year following the intervention, 40% of the children were found to be no longer in need of special education services.”

Torgesen et al. (2001) selected 60 participants between the age of 8 and 10 with severe reading disabilities, gave them 67 hours of intensive phonological awareness and phonics training over eight weeks, plus some followup support, and then tracked them for two years. The results were unexpected and wonderful: the top quartile became strong readers, catching up and even passing their ‘normal’ peers. About 40% of the cohort returned to regular classrooms. Not everyone succeeded – the bottom quartile remained in special-education classrooms and continued to fall behind.

Torgesen’s students had all been identified as learning-disabled and were in LD classrooms. About 30% of students are considered ‘at risk’, and poor readers usually receive some attention before they get so far behind that they are labelled with an LD.  Reading Recovery is typical of the widely-used non-intensive programs for the bottom 20% of at-risk readers.  Typically, RR specialist teachers in reading rooms provide 30 minutes of daily one-on-one coaching.  This works for most students, but 30% of enrolled students usually do not complete the program and are referred further down to Special Education (Tunmer 2003).  Reviewing that cascade:  30 out of 100 students are ‘at risk’ poor readers, bottom 6 of 100 get intervention, and worst 2 of 100 into LD classrooms.

That was Torgesen’s target population; the worst readers, roughly 2% of the population. He excluded second-language learners and those with clear medical issues, although many in the study had ADHD and some were being medicated. They looked like healthy, normal kids; kids you could not identify from watching in the schoolyard. But they had already failed; they were labeled with learning disabilities, they were the ‘dumb kids’, unable to read, barely teachable, no hope, no future. They were warehoused in LD classrooms while their age peers were busily making the transition from ‘learning to read’ to ‘reading to learn’.

Stanovich (1986) describes how the cumulative advantage of reading creates a feedback loop that rewards good readers and leaves poor ones on a slower path. These kids had fallen off the train completely, they were left behind sitting on the tracks.  Their future is predictable – they will get passed from grade to grade unable to read a grade-2 chapter book until they hit high-school and drop out.  There are very few legal jobs for illiterate drop-outs.  Many of them will end in prison.

And yet. Torgesen gives these students 8 weeks of intensive remediation – a snap of the fingers – and almost half of them, these worst 2% of readers, bounce back to the mainstream classroom.

It wasn’t because of some magical training program. Torgesen was actually testing two different remediation methods – an early version of LindaMood Bell (ADD) and a home-cooked phonics program (EP). He concluded that both worked equally well, and subsequent studies identify programs that are as good or better. Almost every child improved.  The worst students tended to improve the most, and even students with the least gains showed dramatic improvements. The graph below is reproduced from his paper, showing standard reading scores for a broad cluster of measures.

Torgesen claimed his subjects averaged gains from this short program of FOUR TIMES more than from the previous 16 months in the LD resource rooms, although that is a complicated thing to calculate.  At the Project, we find that older students gain about a year in reading ability every month with intensive instruction, and other reading specialists have corroborated that.

The Students Who Didn’t Succeed

All students thrived in Torgesen’s intense program.  But over the next two years, Torgesen’s results started to widen; only 40% of the students made it back to normal classrooms.

The students that succeeded all had severe learning disabilities, and then suddenly they didn’t.  With a few weeks of help, they had shaken off their reading impairments and caught up with – or even passed – their ‘normal’ peers.  Their success was not predicted by intelligence, remediation scores, race, gender, or subsequent educational opportunities. Torgesen noted only a small effect for older students and those from higher socioeconomic families.

There’s an interesting implication that Torgesen mentions in his paper.  Failed readers are considered to fall into one of two categories:  ‘True Dyslexics’ and ‘Garden-Variety Poor Readers’ (Stanovich, 1988), and both groups have identical symptoms.  The only way to tell them apart is to offer remediation, the Garden Variety students learn to read and the True Dyslexics have something wrong with them that impedes learning. If half of the bottom 2% responded to Torgesen’s remediation, then the incidence of True Dyslexia is AT MOST one percent.  The rest – 29 out of 30 dyslexic kids in our cascade – are just poor readers, there is nothing wrong with them, they need educational help instead of  accommodations or LD classrooms.

… we can estimate that if the interventions used in this study were available to all children with reading disabilities, only about 1 % of children would remain severely impaired in reading accuracy upon leaving elementary school.

We think ALL students in the study would have succeeded if the training had been longer, if the materials had been better, if the training had better suited student’s individual skill deficits, if there was support beyond the short remediation period, support for family literacy, and so on.  And by implication, we think the incidence of True Dyslexia is considerably LESS than one percent.   Dyslexia is a very rare disorder, your child doesn’t have it.

Let me finish with a provocative statement that I will argue repeatedly in forthcoming postings: There is nothing wrong with children who can’t read, the problem does not lie with them, they are completely normal.

Dyslexia is very real and very serious, but it is artificial like obesity.  The causes are in our environment, in our language, in our social, economic, and political incentives, and especially in our defective education system.  And it is very hard to fix once it takes hold.  We know how to teach reading properly.  But we don’t use those techniques, so that’s no comfort to the large fraction of our students who can’t read.