An experimental comparison of additional training in phoneme awareness, letter-sound knowledge and decoding for...

Despite evidence that synthetic phonics teaching has increased reading attainments, a sizable minority of children struggle to acquire phonics skills and teachers lack clear principles for deciding what types of additional support are most beneficial. Synthetic phonics teaches children to read using a decoding strategy to translate letters into sounds and blend them (e.g., c-a-t = "k - ae - t" = "cat"). To use a decoding strategy, children require letter-sound knowledge (LSK) and the ability to blend sound units (phonological awareness; PA). Training on PA has been shown to benefit struggling beginning readers. However, teachers in English primary schools do not routinely check PA. Instead, struggling beginner readers usually receive additional LSK support.

Until now, there has been no systematic comparison of the effectiveness of training on each component of the decoding process. Should additional support for struggling readers focus on improving PA, or on supplementary LSK and/or decoding instruction? We aim to increase understanding of the roles of LSK and PA in children's acquisition of phonics skills and uncover which types of additional training are most likely to be effective for struggling beginner readers.

We will compare training on each of these components, using a carefully controlled experimental design. We will identify reception-age children at risk of reading difficulties (target n = 225) and randomly allocate them to either PA, LSK or decoding (DEC) training. We will test whether training type influences post-test performance on word reading and whether any effects depend on participants' pre-test PA and/or LSK.

Two hundred and twenty-two participants completed the training. Planned analyses showed no effects of condition on word reading. However, exploratory analyses indicated that the advantage of trained over untrained words was significantly greater for the PA and DEC conditions. There was also a significantly greater improvement in PA for the DEC condition. Overall, our findings suggest a potential advantage of training that includes blending skills, particularly when decoding words that had been included in training. Future research is needed to develop a programme of training on blending skills combined with direct vocabulary instruction for struggling beginner readers.