Effects of a randomised reading intervention study: an application of structural equation modelling.

An intensive phonics-based intervention program for nine-year-old Swedish pupils with reading difficulties was performed. Pupils (N = 112) were randomly assigned to either an intervention or a control group. The training was tailored to the Swedish transparent orthography and designed for one-to-one-tutoring during twelve weeks. Previously, reading speed has been shown to be hard to remediate, and one important purpose was to improve reading speed by explicit training. The intervention group showed improvements immediately after intervention in spelling, reading comprehension, reading speed, and phoneme awareness. There were also significant indirect effects from intervention to all variables one year later. Reading comprehension at immediate post-test predicted spelling one year later, and phoneme awareness at post-test predicted both spelling and reading comprehension one year later. The results suggest the importance of a multi-component intervention, even in transparent orthographies, which includes phonics combined with comprehension strategies and fluency training.