In early 2020, I published a blog post about a UK study comparing two interventions for struggling readers. At the time, the only source available was a set of slides sitting on the website of the Nuffield Foundation. Since then, a peer-reviewed paper has been published and on the DDOLL reading forum, Angelita Manzano drew attention to some additions this paper contributes to what was previously known.
The study was a randomised controlled trial that compared an approach known as ‘Structured Word Inquiry’ to an alternative called ‘Motivated Reading’. I have written about Structured Word Inquiry (SWI) before. It is advocated by brothers Peter and Jeffrey Bowers. Peter Bowers delivers training in SWI and Jeffrey Bowers was one of the researchers involved in the randomised controlled trial. Jeffrey Bowers has also published a paper that reanalyses the findings of the 2000 U.S. National Reading Panel, suggesting the evidence for systematic phonics is lacking. In turn, this paper was used as a source in the recent article by Wyse and Bradbury which, with great fanfare, again sought to cast doubt on phonics-based reading instruction.
Structured Word Inquiry is profoundly weird. Its proponents repeatedly insist that reading can be taught, from the very beginning of instruction, by paying attention to morphology, etymology and letter-sound relationships all at the same time, a claim that seems far-fetched and that appears to be at odds with basic cognitive science. Morphology is about the units of meaning that words are compounded from, such as when ‘ed’ on the end of a word denotes that something happened in the past, and etymology is about word origin. Burdening students who cannot sound-out ‘cat’ with these concepts seems perverse.
There is also a question as to the commitment of SWI advocates to teaching letter-sound relationships, known as ‘grapheme-phoneme-correspondences’ or GPCs, given that, “instruction that suggests that the primary purpose of spelling is to represent the sounds of words,” is an apparent violation of one of the main principles of SWI.
Perhaps because of these limitations, when people ask genuine questions such as, ‘How does SWI teach GPCs?’, Peter Bowers has a habit of producing videos such as the one below where the notional student already knows the GPCs ‘ee’ and ‘ea’ that appear to be the target of instruction.
It’s like asking a chef how they made a cake and the chef proceeding to instruct us in how to slice it.
Back to the study. When I first wrote about it, I was puzzled why the researchers compared SWI to Motivated Reading, a programme I had not heard of. I could have perhaps picked this up back in 2020, but the peer-reviewed article makes it plain that Motivated Reading was created for the purposes of this study. The researchers had not picked an off-the-shelf reading intervention to compare with SWI but instead, they had used a programme that, in principle, nobody else uses - a quintessential straw man. The fact that SWI was no more effective than a programme that was invented purely to test against SWI seems highly significant.
However, the main question is why the researchers did not compare SWI to a systematic phonics approach. After all, systematic phonics is what Jeffrey Bowers had in his sights when he criticised the National Reading Panel. It is systematic phonics that has been adopted in the UK and it is systematic phonics that inspired the phonics screening check, an ongoing bugbear of phonics sceptics. So why, when there was a chance to compare SWI to systematic phonics and potentially slay this dragon, did the researchers miss the shot? Enter a mysterious ‘phonics expert’ who is mentioned in the footnotes of the peer-reviewed study:
“In our initial funding application, the intention was to compare SWI with a phonics condition. However, when planning the interventions, we consulted with a phonics expert, who had concerns about delivering small-group phonics intervention to mixed-ability groups of students in grades 3 and 5, and advised that best practice at that age would be to tailor intervention to the needs of individual students. Unfortunately, this was not feasible within our research design, in which students were randomly assigned to receive intervention in small groups (i.e., neither individual instruction nor ability grouping was possible). Therefore, we developed the MR condition to ensure a stringent test of the effectiveness of SWI.”
How strange. The subjects were all struggling readers, otherwise they would not have been in a reading intervention, so this use of the term ‘mixed ability’ is an unusual one. It also causes me to wonder whether those who initially funded the research were all good with this switch.
Finally, the materials used in the study are posted online so you can decide for yourself why, “teaching assistants found Structured Word Inquiry instruction challenging to deliver.”
This sort of thing would be shocking if it weren't par for the course for so much of educational research.