Paper presented at BERA 1996 conference: University of Lancaster
The Children’s Reading Choices Project
Martin Coles & Christine Hall
University of Nottingham
Abstract:
In the early 1970s The Schools Council in England conducted an inquiry which sought to discover the extent and kind of children’s voluntary reading, what satisfactions they seek from various texts, and what environmental factors influence their choice. This paper draws on a replication of the 1970s investigation and reports findings from a national survey conducted with some 8000 children of 10, 12, and 14 years old in 170 schools using a random stratified sample. The paper will report statistical data on the amount and type of children’s book reading in the U.K. correlated to age, sex, ethnicity, and socio-economic group, and then offers an analysis of these findings.
Key words: national survey children’s reading
In 1971 in England the Schools Council commissioned an enquiry which sought to discover the extent and kind of children’s voluntary reading, what satisfactions they seek from books and magazines, and what environmental factors influence their choice. A national questionnaire survey was conducted by the University of Sheffield under the direction of Frank Whitehead (Whitehead 1977) with some 8000 children of 10, 12, and 14 years old. To supplement the written questionnaire, follow-up interviews were held with 576 children in 34 different schools throughout the country.
It is now over twenty years since we acquired detailed information on children’s voluntary reading habits. In those two decades there have been dramatic shifts in children’s leisure habits and interests and in the cultural climate generally. Such things as changes in the quantity and content of television viewing , the advent of microcomputers, and the marketing of music and fashion for younger children made it seem likely that the 1971 findings did not relate to today’s children. For instance in 1971 the named favourite writer of all children in each age- and sex-grouping was Enid Blyton, and the favourite four periodicals/comics were Dandy, Beano, Buntie and Jackie. The most widely read book at ages 12 and 14 was Little Women.
One of the recommendations of the 1971 survey was that similar surveys should be carried out at ten year intervals, and it seemed important to us, especially in the light of current political concerns about reading generally, that monitoring of children’s reading habits should continue. We set out therefore to replicate the Whitehead study in the context of the 1990s.
This paper reports the findings of that project, the W H Smith Children’s Reading Choices Project, the largest in-depth study of children’s voluntary reading in the U.K. since Whitehead’s. It is divided into three parts. We will very briefly give an indication of the methods used to collect and analyse our data. Then summarise the main findings, especially in relation to change over time, and finally and again briefly, offer some suggestions about the implications of our study for teachers and for further research.
The Project Research Procedures:
A questionnaire was constructed to mirror Whitehead’s investigation insofar as this was possible. Whitehead’s questionnaire is very well designed to probe deeply into children’s voluntary reading habits and it forms the base of our questionnaire. This replication allowed us to gather data which would enable analysis of change over time.
The sample nature needed to replicate, as far as possible, that of the original survey. The strategy we adopted for school selection consisted of sampling from selected local authorities, maintaining representativeness by school size. The LEAs were chosen to span the main socio-economic regions typically recognised in educational and social research. The dominant descriptors here are rural, metropolitan and London. Fifteen LEAs were ultimately selected.
Our selection of pupils was simplified considerably by using intact classes. The relevant ages determined which school year was to be involved. The classes themselves were tutor groups rather than subject groups in order to ensure any possible ability or sex bias was avoided. We are confident in other words that our stratified random sample is so selected that the findings can be generalised to the total relevant populations of children in England.
Over 9,000 questionnaires were distributed. The return rate was 89% giving a total return of 7,976 (10 year olds = 2,975, 12 year olds = 2,455, 14 year olds = 2,546).
A follow-up interview phase of the project aimed to interview 1% of the total sample: in the event 87 children were interviewed (just over 1% of the sample)randomly selected in schools from different regions of England according to the divisions that we had operated with in the construction of the original sample.One interviewer conducted all of the pupil interviews to maintain reliability, and to allow development of ideas and investigations over the range of interviews.
In the analysis of the questionnaires various pupil indicators, especially sex, age, socio-economic and ethnic background, presented a major focus. Obviously the last two categories are problematic. Socio-economic groupings here are labelled A, B, C1, C2, D/E and unclassified in accordance with the Registrar General’s classification, which is based on the occupation of head of household. Children themselves designated the occupation of the adults they live with and clearly those designations are open to interpretation in terms of the categories. The normal caveats which would apply to any categorisation of people into socio-economic groups, need to be applied even more carefully in this case.
When considering the children’s ethnicity, it is difficult to arrive at clear information and definitions. Children’s self report on this necessarily openly phrased question means that the data may be unreliable, in an area of investigation which is anyway open to debate and discussion. In fact 4.5% of the sample did not provide information on ethnicity. The categories we worked with were agreed with the Racial Equality Council.
It is important to emphasize that in this particular survey statistical significance could by itself be used as an indicator of educational significance. The very large sample necessarily leads to an excessively high number of significant tables (when tested using chi-square) within the normal conventions of probability levels of .05 and .01. Even with a more stringent level (p < .001) some judgement is needed about the social significance of a finding.
Main findings:
The ‘headline’ findings then are these:
1. Over the last two decades reported book reading has increased for 10 year olds of both sexes, and for 12 year old girls. It has remained at the same level for 12 year old boys and for 14 year old girls. It has declined for 14 year old boys.
The following tables show the 1971 and 1994 findings on the average numbers of books read in the four weeks prior to the survey.
Table i. Average number of books read by 10+ children , by sex
10+ boys 1971 / 10+ boys 1994 / 10+ girls 1971 / 10+ girls 19942.69 / 2.98 / 3.28 / 3.71
Table ii. Average number of books read by 12+ children, by sex
12+ boys 1971 / 12+ boys 1994 / 12+ girls 1971 / 12+ girls 19941.99 / 1.90 / 2.48 / 2.93
Table iii. Average number of books read by 14+ children, by sex
14+ boys 1971 / 14+ boys 1994 / 14+ girls 1971 / 14+ girls 19941.77 / 1.45 / 2.15 / 2.06
A ‘T-test’ was conducted to test the significance of the differences between the mean scores. There are highly significant differences for 10 year old boys, and 10 and 12 year old girls. In each of these cases the number of books read has increased in the two decades between surveys. There is also a highly significant difference in the figures for 14 year old boys; here the number of books read has decreased. Differences in the figures for 12 year old boys and 14 year old girls are not statistically significant.
2. The average number of books reported as read by children in the four weeks prior to the survey was 2.52 which compares with a figure of 2.40 in the 1971 survey
Up to thirteen spaces were allowed on the questionnaire in which children could record the books they had read over the past four weeks. Of the 7976 children who responded, only three cited thirteen books, filling all the spaces. The average number cited was 2.52.This figure does not, of course, take account of the length of the books.
3. Most children report regular reading, but there is a tendency towards fewer books being read as children get older.
When asked whether they had read any books in the month prior to the survey, 91% of 10 year olds responded positively ( 87.4% in 1971). This compared with 81% of 12 year olds (71.75% in 1971) and 64% of 14 year olds (63.8% in 1971).
4. More girls than boys read books regularly.
More girls than boys had read a book in the month prior to the survey - 84% of girls and 75% of boys. However, avid readers -i.e. those who read a large number of books - can be found in both sexes.
5. There is a clear pattern of decline in the amount of book reading from higher to lower socio-economic groups.
88% of children from the socio-economic group A had read a book in the month before the survey; this figure declines gradually to 76% of socio-economic group D/E. The average number of books read also declined from higher to lower socio-economic groups, from 2.92 in socio-economic group A to 2.40 in group D/E.
6. Children have very eclectic reading tastes.
Returns from the survey resulted in a booklist of 19,344 entries, a large number of which were repeated titles. This list was then re-ordered alphabetically by title and author, and for the different age and sex groups, to discover which were the most popular individual titles, authors and series. We were impressed by how many children were able to ascribe authorship accurately.
It immediately struck us that the children in our survey have very eclectic reading tastes. For instance, it is not uncommon to find respondents who are reading both Virginia Andrews and Charlotte Bronte or Jane Austen novels. One respondent had read Stephen King’s Misery with Roald Dahl’s George’s Marvellous Medicine . One girl’s reading in the previous month consisted of Enid Blyton’s The Naughtiest Girl in the School, Cinderella , Having A Baby , and A Beginner’s Guide to Feminism. One respondent’s reading diet includes Charlie and the Chocolate Factory, and The Rabbit’s New Home, with the Point Horror title Babysitter III. Another’s includes Macbeth, A Midsummer Night’s Dream, Diary of an Anorexic Girl, and Michelle Magorian’s story Goodnight Mr Tom.
For both boys and girls this eclecticism increases with age. 14 year old boys read a wider range of books than 12 year olds, who in turn read a wider range than 10 year olds. This is also true for girls. However, series of books are especially popular with girls and many girls are reading the same series, particularly Point Horror. This complicates the picture. It would seem that there are two trends. In general girls read are reading an increasing range of books as they get older. Those books that are read by large number of girls mainly occur in series.
7. Children of any one age are reading books of markedly different levels of sophistication.
The range of books being read by children of the same age is notable. For example, in one class the leisure reading of one girl is Cinderella while another is reading Nick Fisher’s Pocket Guide to Sex. Some pupils in the same class are reading Enid Blyton and The Animals of Farthing Wood at home, while others are reading Stephen King’s Pet Cemetery. It is noticeable, therefore, not only that individual tastes are eclectic, but that children of the same age are reading books with very marked differences in levels of sophistication in terms of language, plot and subject matter.
These individual profiles struck us as interesting and worth investigating further. In order to do this, we categorised the readers themselves into five broad types:
- reader solely of children’s fiction;
- reader solely of adult fiction;
- reader solely of non-fiction;
- ‘hybrid’ reader, ie a reader of books from more than one category;
- unclassifiable, including children who had not read a book in the previous four weeks.
Obviously there are no hard and fast distinctions to be drawn between ‘children’s fiction’ and ‘adult fiction’, or, indeed, between fiction and non-fiction. Nevertheless, this broad categorising of reader type has allowed useful distinctions to be drawn. In deciding upon the children’s or adults’ fiction categories, we considered where possible the marketing of the books and what that indicated about the expected readership. Books marketed for younger teenage children were put in the children’s fiction category; books marketed for older teenagers as well as adults (eg Terry Pratchett’s or Sue Townsend’s books) were put into the adult fiction category. We included the distinction between fiction and non-fiction readers here because we wanted to consider the percentage of children who read only one or the other, particularly in relation to the often quoted idea that many boys tend to read only non-fiction.
Categorised in this way, over 50% of the sample are readers of children’s and teenage fiction. A sizeable proportion (10%) of children read only what might be considered ‘adult’ fiction. More of these readers come from socio-economic groups A and B than other groups. It is also noticeable that ‘adult’ fiction is more popular with boys than girls at 12+, although by 14+ both sexes are enjoying books that might be enjoyed by their parents.Among these are books by Agatha Christie, Stephen King, Terry Pratchett, Tom Clancy, Virginia Andrews, John le Carre, and Alistair Maclean. The themes of romance, growing up and horror and relationships which figure importantly in children’s reading choices at all of the ages appear in these choices in more sophisticated forms e.g To Kill a Mockingbird, Wuthering Heights, Jane Eyre, Little Women, Silence of the Lambs, Pet Cemetery, Dracula, Misery.
The figure of 16% of the children who are reading across the categories - both ‘children’s’ and ‘adult’s’ fiction and non-fiction - is an indication of the range and variety of children’s reading patterns which struck us so forcibly as we read through their questionnaires.
8. Roald Dahl’s books are overwhelmingly the most popular individual titles.
Roald Dahl’s The BFG tops the list of the most often cited book title, with 267 mentions. This is followed by Matilda, The Witches and The Twits . Charlie and the Chocolate Factory is in sixth place with 197 mentions. Seven of the top ten titles, and ten of the top fifteen, mentioned are by Roald Dahl. This is even more notable since the books referred to are individual titles, whereas other popular titles e.g.The Famous Five and Adrian Mole refer to series of books.
Children were asked whether they have a favourite writer or series of books. Of the 7976 respondents, 65% did name a favourite. 23% of the total sample named Roald Dahl as their favourite author - three times as many as named the second favourite on the list.
Overall, 24% of boys and 21% of girls choose Dahl as their favourite, however boys and girls tend to like different books that he has written; it is notable, for example, that The Twits is a particular favourite amongst the boys but not amongst the girls. Dahl’s popularity declines somewhat as children grow older: 30% of 10 year olds name him as their favourite, 26% of 12 year olds and 12% of 14 year olds.
9. Children’s ‘classics’ are still being widely read.
Books which might be considered as children’s ‘classics’ are being widely read: The Hobbit, The Secret Garden, The Lion, The Witch and the Wardrobe, and Black Beauty all appear in the top twenty most read titles. Lord of the Rings, The Wind in the Willows, Little Women, Alice in Wonderland, The Railway Children, Oliver Twist, The Diary of Anne Frank, Heidi, Treasure Island, are all in the top one hundred titles. Winnie the Pooh, Jungle Book, and Peter Pan also appear. It is likely that sometimes the ‘Disney’ versions are being read, though the original authors are regularly ascribed, which provides evidence to suggest that the originals are also being read.
In the next most popular one hundred titles other ‘classic’ titles appear: White Fang, Wuthering Heights, Jane Eyre, Animal Farm, Around the World in Eighty Days, Gulliver’s Travels, Great Expectations, Kidnapped, To Kill a Mockingbird, David Copperfield, The Magician’s Nephew, Pride and Prejudice, The Horse and His Boy, The Hound of the Baskervilles. One in six titles in our top two hundred then might be considered children’s ‘classics’ and this does not take into account titles which many might consider ‘modern classics’ such as The Borrowers, The Iron Man, Watership Down, Charlotte’s Web, Stig of the Dump, and The Silver Sword.
10. Children enjoy reading series of books. Enid Blyton stories and the Point Horror series are the most popular.