I asked New South Wales’ Centre for Education Statistics and Evaluation (CESE) for some data on the proportion of government schools that are still using Reading Recovery. I had seen a figure of around 60% in the press but wondered whether this was up-to-date, especially since a number of reports have recently been released. It looks like it is spot-on:
“As at February 2016, approximately 1000 NSW government schools with primary-aged student enrolments implement Reading Recovery. This represents a percentage of approximately 60% of eligible schools (ie Primary and Central Schools). Please note: This figure does not include Catholic & Independent schools that use Reading Recovery”
Why does this matter?
Reading Recovery is an expensive intervention because it involves one-to-one tuition. It evolved out of a ‘whole language’ approach to teaching reading. We now know that whole language is flawed, with various national reports finding in favour of systematic synthetic phonics and an explicit approach to teaching letter-sound relationships (see here, here and here). The 2006 Rose report from the UK made a specific point of criticising the idea of multi-cuing, a tactic that is employed in Reading Recovery. Rose suggested that it is potentially harmful because it discourages the proper decoding of words.
Yet the evidence seems to support Reading Recovery. A brief search will return randomised controlled trials (RCTs) that find a positive effect in favour of it and RCTs are supposed to be the gold standard of evidence. The problem with many of these trials is that they don’t test Reading Recovery against an alternative plausible intervention such as Multilit. Instead, Reading Recovery tends to be compared with no intervention at all. Small wonder that children who are given an additional one-to-one reading lesson with an adult make more progress than those who are not, whatever the quality of the intervention. And we might expect even more of a boost if the standard classroom instruction is something like whole language which we know to be less effective. The RCTs do indeed show, to a high standard of proof, that a Reading Recovery intervention is better than doing nothing.
Those who have studied Reading Recovery more closely finding it wanting. James Chapman is a professor at Massey University in New Zealand, the country that is the home of Reading Recovery:
“New Zealand research shows that at best, children who make some progress as a result of Reading Recovery tend to lose the gains after a few years. At worst, our longitudinal study at Massey University showed that children who were said to be successful in Reading Recovery were still, on average, one year behind their same age peers 12 months after completing the programme.”
CESE recently completed its own study in New South Wales. The findings are summarised in the CESE newsletter:
“The results showed some evidence that RR has a modest short-term effect on reading skills among the lowest performing students. However, RR does not appear to be an effective intervention for students that begin Year 1 with more proficient literacy skills. In the longer-term, there was no evidence of any positive effects of RR on students’ reading performance in Year 3.”
So it doesn’t work in theory and it doesn’t work particularly well in practice, even in New South Wales. Rational people might conclude that Reading Recovery should be dropped in favour of cheaper, small-group, synthetic phonics instruction of the kind that the various national reports recommend.
But 1000 schools? That’s a pretty big sunk cost. Nobody will want to admit to being wrong on such a large scale and so I predict that Reading Recovery will continue for a little while yet out of sheer stubbornness if nothing else. And giving the glowing results from RCTs in the U.S., we might even see a resurgence there.