We are testing a new beta website for gov.scot go to new site

The Effects of Synthetic Phonics Teaching on Reading and Spelling Attainment

Listen

THE EFFECTS OF SYNTHETIC PHONICS TEACHING ON READING AND SPELLING ATTAINMENT

CHAPTER THREE PRIMARY 1

COMPARISON OF THE EFFECTIVENESS OF ANALYTIC AND SYNTHETIC PHONICS TEACHING

3.1 In the present study we wished to investigate whether analytic phonics teaching would be found to be as effective in developing reading and spelling skills as synthetic phonics if there was an additional phonological awareness training programme.

3.2 Altogether we studied 304 children in 13 Primary 1 classes in Clackmannanshire. Our interventions began shortly after the children started school at around the age of 5. We had three teaching programmes for the class teachers to implement.

ANALYTIC PHONICS ONLY GROUP

3.3 Four classes were taught about the relationship between letters and sounds using an analytic phonics approach ( see Chapter 1).

ANALYTIC PHONICS+ PHONOLOGICAL AWARENESS GROUP

3.4 Four classes carried out a programme where in addition to analytic phonics teaching, children were taught how to segment and blend spoken words at the level of both rhymes and phonemes, without the aid of print or letters.

SYNTHETIC PHONICS GROUP

3.5 Five classes of children were taught by a synthetic phonics approach ( see Chapter 1).

3.6 The programmes lasted for 16 weeks, the children receiving their interventions via scripted whole class programmes which lasted for 20 minutes a day.

3.7 The children in each of the three groups were assigned the deprivation scores that Clackmannanshire Council had calculated for their school. Using the Council's classification of these scores into advantaged and disadvantaged, it was found that the groups differed in degree of deprivation, F(2,299)=140.8, p<.001. Newman Keuls tests showed that the synthetic phonics group were more deprived than the other two groups, who did not differ from each other. It was important to examine the effects of socio-economic status, as significant social class differences in reading have been found. In the National Child Development Study (Davie et al, 1972), children whose fathers' had unskilled occupations had 5 times the reading problems of children whose fathers' had professional or managerial jobs, at the age of 7. More recently, Stuart, Dixon, Masterson and Quinlan (1998) have shown that in Reception, Year 1, and Year 2 classes in London, the children with middle class parents had significantly better word reading skills than those with working class parents. Finally, Duncan and Seymour (2000) have found that children in Primary 1 in Scotland show social class differences in word reading ability, children from better off homes reading better. Thus as the synthetic phonics group was more deprived than the other 2 groups, if it showed better literacy skills, then this would be the opposite of what would be predicted from the socio-economic backgrounds.

RESULTS

3.8 At pre-test, see Table 3.1 for means and standard deviations, the children in the three groups were found to be matched on all tasks except for knowledge of letter sounds

F(2,301)= 3.3, p<.04; the analytic phonics only group knew more letter sounds than the other two groups.

3.9 At the first post test, see Table 3.1 for means and standard deviations, it was found that the groups differed in single word reading ability (British Ability Scales Word Reading Test, Elliott et al 1977), F(2,289)= 30.7, p<.001; Newman Keuls tests showed that the synthetic phonics group children had significantly higher reading ages than the other two groups, who did not differ from each other. The groups also differed on the more sensitive test of emergent reading (Clay, 1979), F(2,289)= 27.2, p<.001; Newman Keuls tests showed that not only did the synthetic phonics group children perform better than the other two groups, but the analytic phonics only group performed better than the analytic phonics+phonological awareness group. There was a group difference in nonword reading, F(2,289)= 57.8, p<.001, Newman Keuls tests showed that nonword reading was better in the synthetic phonics group than the other two groups, and that the other two groups did not differ from each other. There was also a group difference in the ability to spell dictated words (Schonell & Schonell, 1952), F(2,289)= 57.7, p<.001; Newman Keuls tests showed that the synthetic phonics group had higher spelling ages than the other two groups, who did not differ from each other. Knowledge of letter sounds was also differentially affected by the training schemes, F(2,289)= 74.2, p<.001; Newman Keuls tests showed that the synthetic phonics group were ahead of the other two groups, although at the pre-test they had been behind the analytic phonics group.There was also a group difference in phoneme segmentation skill as measured by the Yopp-Singer Test (Yopp, 1988), F(2,289)= 57.1, p<.001;Newman Keuls tests showed that although the analytic phonics+ phonological awareness group was significantly better at this task than the analytic phonics only group, both of these groups were outperformed by the synthetic phonics group. The groups differed in their ability to produce rhymes for auditorily presented words, F(2,289)=6.8, p<.001 Finally, in terms of irregular word reading, the groups were found to differ, F(2, 289)=10.3, p<.001. Newman Keuls tests showed that the synthetic phonics group outperformed the analytic phonics only group, but not the analytic phonics+phonological awareness group; the two analytic phonics groups did not differ from each other. See Table 2 for means and standard deviations. Newman Keuls tests showed that the synthetic phonics children read these items better than the other two groups, who did not differ. Thus the synthetic phonics trained group outperformed the analytic phonics trained group, despite being from significantly less advantaged homes.

Table 3.1
Mean chronological age, IQ (BPVS), reading age (British Ability Scales Word Reading test), spelling age (Schonell spelling test), emergent reading (Clay Ready to Read Test), letter sound knowledge, phoneme segmentation (Yopp-Singer Test), rhyme skills, and nonword reading (standard deviations in brackets), pre-test and first post-test, Study 2

Research Group

Age

BPVS

Reading Age

Spelling Age

Emergent Reading

Letter Knowledge

Phonemic Segmentation

Rhyme Skills

Nonwords

Pretest

Years

Standardised Score

Years

Years

%

%

%

%

%

Analytic phonics

5.0

92.5

4.9

5.0

0.9

9.0

4.5

17.9

0.3

Controls, n=109

(0.3)

(15.1)

(0.1)

(0.1)

(4.8)

(15.4)

(18.3)

(30.6)

(1.8)

Analytic phonics + phonological

5.0

90.2

4.9

5.0

2.1

3.9

2.7

21.9

0.6

awareness, n=78

(0.3)

(14.0)

(0.4)

(0.1)

(12.5)

(8.8)

(9.9)

(33.1)

(4.6)

Synthetic

5.0

95.2

4.9

5.0

0.7

6.7

4.1

20.0

0.0

Phonics

(0.5)

(16.8)

(0.1)

(0.0)

(6.2)

(14.3)

(14.5)

(29.1)

(0.0)

n=117

First Post-Test

Analytic phonics

5.4

-

5.4

5.2

37.8

58.1

17.2

26.4

8.8

controls

n=104

(0.3)

(0.6)

(0.4)

(24.0)

(24.7)

(27.4)

(36.6)

(22.4)

Analytic phonics + phonological

5.4

-

5.4

5.3

23.9

59.9

34.7

36.4

15.8

awareness, n=75

(0.3)

(0.7)

(0.5)

(25.6)

(24.8)

(44.6)

(36.4)

(29.3)

Synthetic phonics

5.5

-

6.04

6.0

53.4

90.1

64.8

46.5

53.3

n=113

(0.3)

(0.8)

(0.7)

(30.1)

(14.5)

(37.9)

(29.1)

(41.2)

3.10 An examination was also made at the first post-test of ability to read words by analogy. See Table 3.2 for means and standard deviations. The children were asked to read a list of 40 words. They then read 5 clue words that would assist them in reading the 40 words by analogy on second showing, i.e. prior exposure to 'ring' should facilitate the pronunciation of 'sing'. These clue words were then removed, and the 40 words shown again. The items were taken from Muter, Snowling, and Taylor (1994). The gain in reading skill after exposure to the clue words was assessed. It was found on the analogy task that there was an interaction between groups and pre- and post- test reading performance, F(2,289)= 19.1, p<.001. Newman Keuls tests showed that the synthetic phonics children were the only group to show an increase in reading skill between pre- and post-test, and they also showed superior reading to the other groups in both test sessions. There was a significant difference between groups in clue word reading, F(2, 289)= 23.5, p<.001, Newman Keuls tests showing that the synthetic phonics group read the clue words significantly better than the other two groups. Analysis of covariance was therefore used to control for differences in cue reading ability; there was still a significant group difference in gains in word reading at post-test, F(2, 288)= 7.6, p<.001, in favour of the synthetic phonics group. A similar analysis with reading age as the covariate also showed a significant difference between the groups in gain scores F(2,288=8.8, p<.001 .

Thus the ability to read by analogy could not be accounted for in terms of the superior word recognition ability of the synthetic phonics taught group, it indicates a qualitative difference in their approach to reading.

Table 3.2
Mean % correct on Analogy Reading Task and Irregular Word Reading at end of training programme (first post-test), Study 2

Research group

Pre-test scores

Clue word reading scores

Post-test scores

Irregular words

Analytic phonics controls, n=104

2.9
(12.0)

6.3
(18.3)

2.6
(9.3)

21.4
(19.5)

Analytic phonics + phonological awareness, n= 75

4.9
(15.8)

11.4
(27.3)

5.5
(16.2)

15.3
(23.1)

Synthetic phonics, n=113

16.9
(25.7)

30.5
(32.9)

22.7
(23.7)

30.2
(25.4)

3.11 After the first post-test, the two analytic phonics groups carried out the synthetic phonics programme, completing it by the end of their first year at school. The synthetic phonics taught children spent this time consolidating their learning rather than working on further grapheme to phoneme correspondences. In May of the following school year, 15 months after the end of the programme, all of the children were re-tested on the standardised tests of single word reading and spelling. See Table 3.3 for means and standard deviations.

Table 3.3
Mean chronological age, mean word reading age (British Ability Scales Word Reading test), and mean spelling age at second post-test, 15 months after termination of the programme, after all groups had been introduced to synthetic phonics.

Research Group

Age

Reading Age

Spelling Age

Analytic phonics

6.7

7.4

7.5

controls , n=95

(0.3)

(0.9)

(0.7)

Analytic phonics and phonological awareness, n=66

6.7

7.6

7.4

(0.4)

(1.3)

(0.7)

Synthetic phonics

6.8

7.7

7.8

n=103

(0.3)

(1.1)

(0.9)

3.12 There was no longer a significant difference in reading between the three groups, F(2,265) = 2.8, p=.064, although it was close to significance. Separate analyses by sex showed that boys' word reading skills did not differ according to what group they had initially been in, F(2,138)<1, but girls' did, F(2,124)= 4.0, p<.02. Newman Keuls tests for the girls showed that the synthetic phonics group (mean reading age 7.8 years, S.D. 0.9) and analytic phonics plus phonemic awareness training group (mean reading age 7.7 years, S.D. 0.9) both outperformed the analytic phonics only group (mean reading age 7.3 years, S.D. 0.9), and did not differ from each other. In spelling there was a significant difference between groups F(2, 261) =7.4, p<.001. Newman Keuls tests showed that the synthetic phonics group had better spelling than the other two groups, who did not differ from each other. Reading and spelling scores were found to be significantly above chronological age, F(1,267) = 193.2, p<.001, F(1,264)= 337.6, p<.001, respectively. On the test of reading comprehension (Primary Reading Test, France, 1981), there was no significant difference between groups, F(2,255)=2.0, p>.05.

3.13 An examination was also made of whether the children's reading comprehension test scores were higher than the chronological ages; this was found to be the case, F(1,256)= 53.9, p<.001. See Table 3.4 for means and standard deviations.

Table 3.4
Mean chronological age, mean reading comprehension age (Primary Reading Test), second post-test, 15 months after termination of the programme, after all groups had been introduced to synthetic phonics.

Research Group

Age

Reading Age

Analytic phonics

6.8

7.0

controls , n=89

(0.3)

(1.0)

Analytic phonics + phonological

6.7

7.0

awareness, n=64

(0.4)

(1.0)

Synthetic phonics

6.8

7.3

n=105

(0.3)

(1.1)

SUMMARY

3.14 This chapter examined the extent to which children learning by a synthetic phonics approach read and spelt better than children taught either by a standard analytic phonics approach, or by a standard analytic phonics approach supplemented by phonemic awareness training. It was found that:

  • At the end of the experimental programmes, the synthetic phonics group read 7 months ahead of chronological age, and 7 months ahead of the other two groups. They were also 7 months ahead of chronological age in spelling, and spelt 8 to 9 months ahead of the other two groups.
  • At the end of Primary 2, the girls who had had the programme at the start of schooling read better than those initially taught by the standard analytic phonics approach. However, the timing of this programme had no impact on the boys' word reading skills at the end of Primary 2.
  • At the end of Primary 2, the early-taught synthetic phonics group (boys and girls) spelt better than the other two groups