Reading and Writing: An Interdisciplinary Journal 13: 349–355, 2000.
© 2000 Kluwer Academic Publishers. Printed in the Netherlands.
Reading intervention need not be negligible:
Response to Cossu (1999)
PETER J. HATCHER
Department of Psychology, University of York, UK
Abstract. The treatment effects of a reading intervention study by Hatcher, Hulme and Ellis
(1994) have been described as unstable (Cossu 1999) and to be evidence that reading-delayed
children are impervious to intervention. Data are presented to show that, in this study, the
combined reading and phonological awareness training group (R+P) made greater progress in
learning to read than the Reading alone (R) and Phonological awareness alone (P) groups in
addition to that of the unseen control group (C). Group R+P also exhibited effect sizes of 1.1 to
2.6 for accuracy, and 1.3 to 1.6 for comprehension, some nine months after the intervention had
ceased. These data afﬁrm the stability of the treatment effect, the Sound Linkage hypothesis
and the effectiveness of comprehensive and well-structured intervention programmes.
Keywords: Intervention, Phonological awareness, Reading
In his paper ’Biological constraints on literacy acquisition’, Cossu (1999)
speculates that a core component of the brain serves the specialised role
of enabling children to detect the correspondence between orthography and
phonology, and that for reading delayed children the system is relatively
impervious to intervention. As partial evidence for this he suggests that in
the Hatcher, Hulme and Ellis (1994) study there were ’unstable’ effects of
treatment. I shall argue that on the contrary, the effects of the intervention
were substantial and that for reading, they were also enduring.
Summary of Hatcher, Hulme and Ellis (1994)
To summarize this study brieﬂy, 128 7-year-old reading-delayed children
were divided into four matched groups which were randomly assigned to one
of three intervention conditions (reading alone (R), reading with phonology
(R+P) or phonology alone (P)) or to a non-intervention control group (C).
The intervention conditions, which are fully described elsewhere (Hatcher
1994; Hatcher et al., 1995), comprised forty 30-minute individual teaching
sessions, two per week, spread over a 25 week period. The children were
assessed on three occasions, before the intervention (t1), when the interven-
tion was completed (t2) and some nine months after the intervention had
ceased (t3). It was found that, compared to the control group (C), the only