Since our study of synthetic phonics in
Clackmannanshire (Johnston and Watson, 2004) has been widely discussed, though
perhaps not widely read, a remarkable number of myths have been circulating.
Hall (2006, page 12) argues that in our
research in Clackmannanshire ‘analytic phonics was set up for failure’. Actually,
as we are scrupulously careful researchers, our very first study set out to
find out exactly how analytic phonics was taught. Definitions of analytic
phonics are rarely very explicit; below is one of the more detailed ones that
we have found.
analytic method initially presents the student with a list of words that become
part of his sight vocabulary through visual memory techniques. Subsequently he
is taught to analyse these words by identifying certain common sounds that
appear in them. For example, the sight words milk, man and mother are shown to
the child, who is asked if he sees anything about these three words, he then ‘reads’
them. The teacher cautions him to listen to the sound he hears at the beginning
of each word. ‘Whenever we see this letter, “m”, in a word, we will hear the
sound that we hear at the beginning of milk and man. Can you think of some
other words that begin the same as milk and mother?’ The proponents of the analytic
method contend that it keeps decoding as part of the reading act, since the
sounds are never isolated but are always taught within the context of the word
and perhaps the sentence.” Harris, L.A.
and Smith, C.B. (1976)
On pages 10 and 11 of the report on our
longitudinal study (Johnston and Watson, 2005) we described our observations of
classroom practice in 12 classes in Scotland teaching by the analytic
phonics approach, starting in 1992. The programme started by teaching letters
in the initial position of words, at the pace of one letter sound a week. In
all but one class, this took 26 weeks to complete, generally around March of
the first year at school. Children were then taught about the importance of
letters in the final position of words, and then in the middle, at which time
sounding and blending of CVC words could commence. At this point, we found that
independent reading skills really took off. The pace and progression in
analytic phonics teaching that we observed was common practice throughout Scotland at the time (and is similar to how
schools implemented Progression in
Phonics in England).
It was the observation that independent reading skills were very much better in
the class where one teacher took a faster approach, introducing sounding and
blending much earlier, that led us to study synthetic phonics.
In our second study, carried out in 1995,
children learning by the normal classroom analytic phonics approach received
extra phonics tuition for 10 weeks in small groups outside the classroom,
having two 15 minute sessions a week (Johnston and Watson, 2004). One group of
children learnt by the synthetic phonics approach, learning letter sounds at
the pace of two a week. A second group also learnt two letter sounds a week,
following an analytic phonics programme. As this was the first term of school,
these letters were learnt at the beginning of words. In the third group, the
children received the same new print word exposure as the other two groups, but
had no phonics teaching additional to the classroom analytic phonics programme.
By Christmas of the first year at school, the synthetic-phonics-taught children
read significantly better than the other two groups, and also knew more letter
sounds (even though the analytic phonics group had learnt letter sounds at the
same rate). The intervention ended at this point, but the children were
followed up until the start of their second year at school. By this point, the
children had all been taught in their classroom phonics programme to look at letters
in all positions of words (Johnston and Watson, 2004, page 352). Despite the
fact that the classroom programme had reached the sounding and blending stage,
the synthetic-phonics-taught group was reading around 11 months ahead of the
accelerated analytic phonics group. Thus despite the fact that our intervention
in Clackmannanshire (see below) was terminated before the stage where sounding
and blending is used in analytic phonics programmes, our evidence from the
earlier study suggests that the children taught this way were very unlikely to
have been able to catch up with the children taught synthetic phonics.
In our third study of phonics, carried out
in Clackmannanshire starting in 1997, we again compared the effectiveness of
analytic and synthetic phonics teaching (Johnston and Watson, 2004), this time
delivered on a whole-class basis. The children in the analytic phonics
condition followed a systematic scripted daily classroom programme based on the
observations in the first study of how analytic phonics was typically taught.
As the 16-week intervention ended in February of the first year at school, the
children had by that time learnt 16 letter-sounds at the beginning position of
words. As we had established in Study 2 that speed of letter learning in itself
was not a factor in accelerating learning to read, the children in the
synthetic phonics group received a programme delivered at the typical fast pace
of such programmes. They covered letters and letter sequences for the 40+
sounds in the English language in 16 weeks. At the end of the programme, these
children were reading 7 months ahead of the analytic phonics group, and had
much better phonemic awareness skills. However, both groups had been exposed to
the same new print vocabulary.
The children in all conditions read text
from early on, starting on reading scheme books just 6 weeks after the
programme started, contradicting the idea that all synthetic phonics programmes
advocate covering the 40+ sounds before the introduction of reading for meaning.
Much is made of the fact that the synthetic phonics programme in
Clackmannanshire led to much greater increases in word reading and spelling
skill than in reading comprehension, implying that reading comprehension did
not benefit from the intervention. However, it should be noted that at the end
of the seventh year at school, reading comprehension in the study was
significantly above age level, in a sample that had a below average SES
(socio-economic status) profile.
It has been claimed that other interventions
were being carried out in Clackmannanshire at the time of our study, which
could have led to better results for the synthetic phonics condition (Ellis,
2005). At the time of our study, home-school liaison officers were appointed in
four of the study schools. However, two of these were in schools in the
synthetic phonics condition, and two were in an analytic phonics condition, so
these appointments cannot account for the gains found with the synthetic phonics
The DfES-funded Torgerson et al (2006)
review of the effectiveness of reading programmes chose to carry out a
meta-analysis only of studies that had random allocation to conditions. It is
not clear why a study that received so much public funding (around £70,000)
only undertook a partial review of the literature. Like two-thirds of similar
studies in the literature, the Clackmannanshire study did not have random
assignment of schools to conditions. As random allocation was not possible, the
schools from the most disadvantaged areas were allocated to the synthetic phonics
condition. This is a tough test of the effectiveness of synthetic phonics, as
children from areas of deprivation do less well in reading than those from
better-off areas from the very first year at school (Stuart et al, 1998, Duncan
and Seymour, 2000). The National Reading Panel (2000) and Camilli et al. (2003)
reviews looked specifically at whether studies having random assignment to
conditions led to different outcomes from those studies that did not (the
majority of the literature). Neither review found that this made any
difference. Similarly, our second study, carried out in 1995, did have random
allocation of children to conditions, and the Clackmannanshire study replicates
the findings of that earlier study.
The two previous reviews looking at the
effectiveness of reading programmes scrupulously excluded all unpublished
studies. Therefore another singular feature of the Torgerson (2006) review is
that one of the three studies they include that compares analytic and synthetic
phonics conditions is an unpublished study, delivered at a conference in 1971.
This study found no advantage for synthetic phonics teaching. As a peer-reviewed
article has not appeared on this study in 35 years, it suggests there is
something wrong with it. There is indeed a major flaw – the synthetic phonics
programme was not correctly implemented. The kindergarten children were taught
to sound and blend words like 'tape', an approach which would lead to an
incorrect pronunciation of the words. In synthetic phonics schemes such items
are taught much later on, when phonic rules are taught.
Interestingly, the detractors of the
Clackmannanshire study have not attempted themselves to demonstrate that their
preferred method yields as good or better results than a synthetic phonics
programme. Their method seems to be to merely attack the Clackmannanshire study
and thereby imply that the approach that they advocate is as good or better,
without collecting any supportive data.
In sum, any piece of research, particularly
one that has had such a wide influence, should indeed be subjected to close
scrutiny. However, looking at all the points that have been raised about the
study, it is clear that there is a desire in some quarters to denigrate the
work by slur and innuendo, without actually producing any evidence that
contradicts the findings of the studies carried out.
Camilli, G, Vargas, S. and Yurecko, M.
(2003) Teaching children to read: the fragile link between science and federal
education policy, Education Policy
Analysis Archives, 11, No 15, http//epaa.asu.edu.epaa/v11n15/
LG and Seymour PHK (2000) Socio-economic differences in foundation-level
literacy. British Journal of Psychology,
Ellis, S (2005) 'Platform: Phonics' Times
Educational Supplement (Scotland),
Hall, K (2006) How Children Learn to Read and How Phonics Helps. In
Lewis, M and Ellis, S (Eds) Phonics:
Practice, Research and Policy, Sage: London.
Harris, L.A. and Smith, C.B. (1976) Reading Instruction: Diagnostic Teaching in the
Classroom (2nd Edition). Holt, Rinehart and Winston: London.
R.S and Watson, J. (2004) Accelerating the development of reading, spelling and
phonemic awareness. Reading and Writing, 17 (4), 327-357.
Johnston, R.S, and Watson, J. (2005) The
effects of synthetic phonics teaching on reading and spelling attainment, a
seven year longitudinal study. Published by the Scottish Executive Education Department. Available on http://www.scotland.gov.uk/library5/education/sptrs-00.asp
National Institute of Child
Health and Human Development (2000). Report
of the National Reading
Panel. Bethesda, MD: National Institute of Child Health and
Stuart, M, Dixon M, Masterson J, and
Quinlan, P. (1998) Learning
to read at home and at school. British Journal Educational
Psychology, 68, 3-14.
Torgerson, C, Brooks, G, Hall, J (2006) A Systematic Review of the Research Literature
on the Use of Phonics in the Teaching of Reading
and Spelling. Report RR711, DfES: University of Sheffield.