[Spelling Reform, §9.7 pp.146-148]
[Spelling Progress Bulletin, Spring 1980, pp.10-12]
The Effects of a Simplified Spelling in Children's Readiness to Read.
by D. V. Thackray, Ph.D.**This paper was presented at the 2nd International Conference on Reading and Spelling at Nene College, Northampton, England, on July 27-30, 1979. Sponsored by the Simplified Spelling Society.
This report falls fairly naturally into three parts. First I would like to give you the background to the research, then go on to describe the investigation and finally to present the results.
Background to the Research.I have been interested in the field of reading readiness for a number of years. In my first research I tried to determine the relative importance of the generally accepted reading readiness skills, such as visual and auditory discrimination, mental ability and vocabulary development, in learning to read and making progress in reading. Research of this kind has been carried out by American research workers over a period of almost fifty years, from the time reading readiness tests were first published; in England however, to the best of my knowledge, mine was the first experiment of this kind. The reasons for this lack of experimentation in England are firstly, English children begin school when they are five, which is felt to be rather too young an age for widespread testing, and secondly, until recently there were no published British reading readiness tests.
In my first experiment I followed the approach commonly used by American research workers. A representative sample of 183 children was tested in a number of reading readiness skills using an Anglicised version of the American Harrison-Stroud Reading Readiness Profiles (1956), when commencing their second term in school (average age 5 years, 4 months); the children were also tested for three other important factors in reading readiness, namely general ability, home environment, and emotional and personal attitudes. Later when commencing their fourth and fifth terms (average age 6 years; and 6 years, 4 months respectively) the children were given the Southgate Group Reading Test 1 (1959), to measure reading achievement.
The earlier reading readiness results were correlated with the later reading achievement results. And the individual reading readiness skills, which correlated the most highly with reading achievement, were those of visual and auditory discrimination. These correlations were higher than the one for mental age, showing that in this experiment, the readiness skills of visual and auditory discrimination were as important - perhaps more important - than mental age in learning to read in the early stages.
In England, 1961 saw the start of the main i.t.a. experiment under the direction of Prof. John Downing. When describing the differences between i.t.a. and traditional orthography (t.o.), both Pitman (1961) and Downing (1964) have stressed that i.t.a. is simpler both in its visual and auditory characteristics. It is simpler visually because in i.t.a. there is a constant visual pattern for each whole word or sentence; it is simpler from the auditory standpoint because each symbol in i.t.a. stands effectively for its own sound.
Because of its simplicity, protagonists of i.t.a. have suggested that children using i.t.a. should be ready to read at an earlier age than if learning to read with the more complex t.o. Knowing from my first experiment the importance of visual and auditory discrimination, and from the literature that i.t.a. was simpler visually and auditorily, I felt that this hypothesis was a reasonable one and in my second experiment - the one with which this paper is concerned - I decided to test it experimentally.
Purpose of the Research.So the main purpose of my research then was to test the hypothesis that children learning to read with i.t.a. are ready to read at an earlier age than children learning to read with t.o.
The Investigation.The method of approach was to enlist the co-operation of 16 schools; 8 schools where the children were learning to read with i.t.a. and 8 schools, matched as well as possible with the i.t.a. schools, where the children were learning to read with t.o. The original total sample was 300 children with 150 in each group, but family removals and the matching of the two groups reduced these numbers to 119 in each group during the first two years of the experiment and to 102 children in each group during the third year.
The children in the experiment were studied over a three year period, during which time the children learning to read with i.t.a. had transferred to t.o. and had been given the opportunity to make good any setback in reading achievement experienced after transfer. Reading readiness considerations were the main ones in the investigation, but it was realised that true reading standards, needed for comparison with standards on reading readiness measures, are not established until the children who started to read with i.t.a. have been reading for a reasonable length of time in t.o. after the transfer. So this meant testing and observing the children who were taking part in this experiment over a period of three years.
After being in school for approximately six weeks, all the children in the sample were given the Harrison-Stroud reading readiness tests of visual and auditory discrimination, and also tests of visual and auditory discrimination that I constructed. They were also given the W.I.S.C. (1949), and my own test of vocabulary. At the same time, class teachers of the children were asked, firstly, to rate each child on a five point scale for a number of reading readiness evaluations including mental abilities, physical attributes, social and emotional traits and language development; and secondly, to give the fathers' occupations and details of any homes which were other than normal. This information gained from tests, evaluations and teachers' reports enabled the later matching of the i.t.a. and t.o. groups and sub-groups to be made.
At the beginning of the children's third term in school, two of the reading readiness tests, my tests of visual and auditory discrimination, were given to the whole sample. These two tests were given firstly, to measure progress made in these two skills and secondly, to see if the children learning to read with i.t.a. had in any way developed these skills differently from the children learning to read with t.o. This comparison was made because the results of a small experiment carried out by Sister John (1966), suggested that i.t.a. might develop perceptual skills to a greater extent than t.o., and it was decided to test this hypothesis. Also at the same time a first reading achievement test, the Schonell Graded Word Reading Test (1959), was given to all the children. The usual form of the test was given to the t.o. group, but a transliterated version of the same test was given to the i.t.a. group. In this way initial progress in learning to read was assessed.
After a further term, that is at the beginning of the childrens' fourth term in school, the same reading achievement test was repeated together with a second more comprehensive reading test, the Neale Analysis of Reading Ability (1963) ; transliterated versions were used with the i.t.a. children.
Reading achievement and progress was again measured at the beginning of the children's sixth term in school. At this stage, it was found that many children had transferred to t.o. and where this had occurred, the children concerned were tested in t.o. Those children still reading with i.t.a. were tested both in i.t.a. and t.o.; in these cases the t.o. test was given to the children first. As being the more difficult, it was felt that the taking of the t.o. test would not affect the i.t.a. scores to any great extent. A comparison of the i.t.a. and t.o. scores made by the same children, at the same time, on the same test, provided interesting evidence regarding the ease of transfer from i.t.a. to t.o.
The final reading achievement tests of the investigation were given at the beginning of the children's ninth term in school, when some of the children had moved to Junior Schools or Junior Departments, and all but four had transferred to t.o. reading. The same two reading achievement tests were given, but this time only the t.o. versions were used.
Analysis of the Data.In order to compare the reading readiness requirements of children learning to read with i.t.a. and t.o., two groups of children were matched for age, sex, reading readiness skills of visual and auditory discrimination, intelligence, vocabulary and social class. The two matched groups of i.t.a. and t.o. children were then compared in three main ways. Firstly, the mean reading achievement scores of the i.t.a. and t.o. groups were compared throughout this experiment. Table 1 illustrates this approach.
Table 1.showing a comparison between the mean scores of the i.t.a. and t.o. groups on the Schonell Graded Word Reading Test given for the first time (given in i.t.a. to the i.t.a. children; given in t.o. to the t.o. children).
|S.E. of DIFF.||C. R.||STATIS-|
This table is just to illustrate my first approach which was to compare the mean scores of the i.t.a. and t.o. groups on the Reading Achievement Tests given from time to time throughout the three years. Column 1 indicates the two groups; column 2 the number in each group (119) and column 3 - the important column - shows the mean reading achievement score of each group on the Schonell Test given at the end of the first year in school. Column 5 shows the difference in the mean score of 3.25 in favour of i.t.a. The other figures need not delay us, as I am only trying to illustrate my approaches.
Secondly, five levels of performance achieved by sub-groups of i.t.a. and t.o. children on the various reading measures were taken, and for each level the mean scores attained by the sub-groups of i.t.a. and t.o. children were calculated and compared. Table 2 illustrates this approach.
Table 2.showing a comparison of the mean scores attained on the Schonell Graded Word Test, by sub-groups of i.t.a. and t.o. children who attained similar levels of performance on the writer's Visual Discrimination Test.
|Visual Discrimination - Thackray. Schonell Graded Word Reading - first time.|
This table illustrates my second approach which was to compare the mean scores attained on the Reading Achievement Tests by sub-groups of i.t.a. and t.o. children who attained similar levels of performance on the measures of reading readiness skills given soon after the children entered school.
In this particular table, column 1 shows the range of scores possible on my Visual Discrimination Test, divided into 5 levels of performance, 0-6, 7-13, 14-20, 21-27, and 28-34. Column 4 shows the mean reading achievement scores of the i.t.a. and t.o. children who attained similar levels of performance on Visual Discrimination. Column 6 shows the differences in the mean scores of the i.t.a. and t.o. sub-groups and a clear pattern can be seen - the mean scores of the i.t.a. groups are consistently higher than the mean scores of the t.o. groups although they had the same level of performance on the Visual Discrimination Test given initially. From such an approach it is possible to see that i.t.a. children with a lower level of performance in Visual Discrimination than t.o. children could reach the same reading achievement level in the same time. For example with the range of scores 28-34, the t.o. reading score was 4.25 (column 4). If we enter the range of scores 14-20 we see the i.t.a. children's mean reading score was similar (4.96), but this with a lower level of performance in Visual Discrimination. I hope this indicates the way in which I obtained my results.
Thirdly, a comparison was made between the mean scores attained on the reading achievement measures by sub-groups of i.t.a. and t.o. children, with similar mental ages. Table 3 illustrates this approach.
Table 3.showing a comparison between the mean scores attained on the Schonell Graded Word Reading Test, given the first time, by sub-groups of i.t.a. and t.o. children with similar mental ages.
|No. of i.t.a.
chn. in each
mental age grp.
|No. of t.o. children
in each mental
|Mean score of
|Mean score of
This table illustrates my third approach which was to compare the mean scores attained on the reading achievement tests by sub-groups of i.t.a. and t.o. children with similar mental ages.
Across the top of the table you see eight mental age ranges from below 3 years, 6 months to 6 years, 11 months.
If you look at the column headed 4 years, 6 months to 4 years, 11 months, you see 23 i.t.a. children fell into this mental age range, and 20 t.o. children fell into this range. The mean scores of the i.t.a. children in the group was 3.69 and the mean score of the t.o. group was 1.65. This is a common pattern indicating that with similar mental age levels i.t.a. children score consistently higher than the t.o. children, and it follows that with lower mental age levels, i.t.a. children can score the same as the t.o. children.
Main Findings.1. In my sample, i.t.a. had no more favourable effects on the growth of perceptual discrimination skills than had t.o. so Sister John's earlier findings were not borne out.
2. Regarding the first statistical approach in which mean reading scores of the matched groups were compared throughout the experiment, the following results were established:
i) When the i.t.a. group was tested in i.t.a., there were significant differences between the mean scores of the i.t.a. and t.o. groups, in favour of i.t.a. As the two groups were well matched, the children in my sample learned to read more easily and made better progress with i.t.a. than with t.o. Conversely, the traditional alphabet and spelling of English used with an eclectic approach was a more difficult medium for the teaching of reading than i.t.a.
ii) When the two groups were tested in t.o. at the end of their second and third years in school, there were no significant differences between the mean scores of the i.t.a. and t.o. groups. When i.t.a. children read in the relatively more difficult medium of t.o., the average score was lowered and the i.t.a. group lost its early lead.
iii) At the end of the second year, a comparison was made between the mean scores attained on the i.t.a. and t.o. versions of the two reading achievement tests by 50 i.t.a. children who had not transferred to t.o. There was a highly significant difference between the mean scores on the i.t.a. and t.o. versions of both tests, indicating that for these 50 children at this stage, the t.o. version of the test was much more difficult for them to read than the i.t.a. version and again shows that in my experiment there was a setback in reading progress during the transfer stage.
3. Regarding the second statistical approach which compared the mean reading achievement scores of sub-groups of i.t.a. and t.o. children who attained similar levels of performance on the reading readiness measures given initially, the following results were established:
i) When the i.t.a. group was tested in i.t.a., the results show that for nearly all levels of performance on the reading readiness tests, the mean reading achievement scores attained by the i.t.a. sub-groups are greater than the mean reading achievement scores attained by the t.o. sub-groups and in many cases significantly greater. This pattern of results indicates that i.t.a. sub-groups with lower levels of reading readiness than t.o. sub-groups can reach similar levels of reading achievement to those t.o. sub-groups, whilst reading in i.t.a. If i.t.a. children can learn to read with lower levels of reading readiness than t.o. children, then i.t.a. children, on average, will be ready to read earlier than t.o. children.
ii) When the two groups were tested in t.o. at the end of their second and third years in school, and a comparison again made of the mean reading scores of i.t.a. and t.o. sub-groups who attained similar levels of performance on the reading readiness measures given initially, a new pattern of results emerged. The mean reading scores of the sub-groups were similar, again providing evidence of the setback in the progress of i.t.a. children at the transition stage.
4. Regarding the third statistical approach which compared the mean reading achievement scores of sub-groups of i.t.a. and t.o. children with similar mental ages initially, the following results were established:
i) When the i.t.a. group was tested in i.t.a., the figures indicated that i.t.a. children were able to learn to read as well as t.o. children with an average mental age of six months to a year less than the average mental age of the t.o. children.
ii) When both groups were tested in t.o., the results indicated that the i.t.a. and t.o. sub-groups with similar levels of mental ability initially had similar levels of reading ability, again providing evidence of the setback in the reading progress of i.t.a. children during the transition stage.
I feel that my research showed experimentally that:
a.) i.t.a. is simpler than t.o. in its visual and auditory structure;
b.) i.t.a. children are ready to read earlier and make quicker progress than t.o. children taught with an eclectic approach;
c.) there is a setback for the i.t.a. children during the transfer stage which resulted in similar mean reading scores for the i.t.a. and t.o. groups at the end of three years in school.
Conclusion.If firstly, children learning to read with i.t.a. were taught with confidence at a rather earlier age than is normal for the teaching of reading with t.o., and secondly, the transfer to t.o. could be made easier in some way, then i.t.a. children could keep their lead and reading standards could be raised.
In the discussion of i.t.a in the Bullock Report (1975), the Committee made the following two comments, which are relevant to this paper:
" ... we have already noted the bewildering complexities of the English spelling system, and it is self-evident that a simplification of the relationship between sound and spellings must make it easier for a child to make progress in the early stages. If there are fewer items to be learned this alone must reduce the time required, and if there are fewer ambiguities there will be less confusion. All this is amply confirmed by research."
"As a Committee we are not unanimous on the value of i.t.a. but we believe that as there is no evidence of adverse side effects at a later stage, schools which choose to adopt it should be given every support. We also feel that teachers should examine the question of i.t.a. on its merits."
The Bullock Committee is encouraging teachers to look again objectively at i.t.a., and I would endorse this view.
Downing, J.A. (1964), The i.t.a. Reading Experiment. London: Evans Brothers.
DES (1975), A Language for Life (Bullock Report), HMSO.
Harrison, M.L. and Stroud, J.B. (1956), The Harrison-Stroud Reading Readiness Profiles. Houghton Mifflin Co.
Neale, M.D. (1963), Neale Analysis of Reading Ability, London: MacMillan Ltd.
Pitman, I.J. (1961), Learning to Read: An Experiment, Journal of Royal Society of Arts. 109. pp. 149-180.
Schonell, F.J. and Schonell, F.E. (1950), A Graded Word Reading Test. Edinburgh: Oliver and Boyd.
Sister John, "The Effect of the i.t.a. medium on the Development of Visual and Auditory Awareness of Symbol Differences", in Downing, J.A. The First International Reading Symposium. London: Cassell. pp. 112-123.
Southgate, V. (1959), Southgate Group Reading Test 1. University of London Press.
Wechsler, D. (1949), Manual for the Wechsler Intelligence Scale for Children. New York: Psychological Corp.
Back to the top.