THE IDENTIFICATION OF THE VARIABLES
2.2 The variables in the miscue analysis
2.2.1 The development of miscue analysis
The main part of the research is concerned with the analysis of miscue variables made when reading. Of relevance is some reference, in these two sections, to the work of Kenneth Goodman in the United States and to the development of miscue analysis in Britain.
In assessing children's oral language and written work, teachers in primary schools may set their own standards, often based on prior experience and consultation with other staff. This contrasts with the range of standar dised tests which may be used to assess reading ability. These tests, standardised for use with indigenous pupils, reflect a cultural perspective that disadvantages bilingual children. However, a diagnostic assessment can be made of children' s reading skills and the extent of the children's mastery of phonic rules and comprehension through the use of "miscue analysis".
"Miscue analysis" was first developed by Kenneth Goodman (Gollasch, 1982). His investigation into the reading miscues or errors that children make when reading was undertaken for the purpose of providing knowledge of the "reading process" (Goodman, 1973). This knowledge, he suggested, can form the basis for more effective reading instruction. In trying to establish a model of the reading process, he found he was mainly concerned with the miscues or errors that children made when reading and that these were of greater significance than the parts of the text that the children could read accurately. He also observed that when readers make reading errors they are not random errors and that therefore they are capable of analysis (Goodman, 1969).
Goodman (1973) defines miscues in reading as "an actual observed response in oral reading which does not match the expected response." For each error, a linguistic cause can be identified. The errors or miscues made when reading are not accidental and can be related not only to what is on the printed page but ,also to the linguistic knowledge the reader brings to the text. The miscues arise as a result of the child attempting to process the print in order to reach meaning and, for each miscue, there will be one or more identifiable linguistic causes. The child is using prior experiences and knowledge of language to decipher what is on the printed page.
Each miscue, therefore, can be considered on a number of variables so that the researcher considers the linguistic
information inherent in the error. Each error might be considered for its relationship to the visual input, for its proximity to the grammatical structure within which it is contained and for the degree of approximation to the meaning of the text (Gollasch, 1982). Any miscues made will be diagnostically useful for the class teacher, as they will indicate the extent to which the students are using grapho-phonic, syntactic and semantic know ledge. Teachers can then identify the strengths of the children and where they need help to become more effective readers.
Because of the complexity of Goodman's work in its original form, it had little practical classroom use, as making an analysis of each miscue was time consuming. Goodman and Burke (1973), for example, listed eighteen questions that could be asked of each substitution error. In Britain, adaptations have been made to Goodman's system of miscue analysis that make the coding simpler and are of more general use for the teacher. In the main, however, the principles behind Goodman's theory have been kept intact.
2.2.2 Miscue analysis in Britain
The Neale Analysis of Reading Ability (1958) was one of the first diagnostic tests to categorise the variables or types of errors made when reading, dividing them into substitutions, omissions, reversals and non-responses. By 1970, Biemuller, tracing the reading development of six to seven year olds, concluded that "miscue analysis
was a powerful diagnostic tool for the assessment of reading." Pumfrey (1977) also suggested that miscue analysis might be used as an indication of progress in reading. Each miscue needed to be examined for the extent to which preceding and succeeding context had been used. Dean et al (1978) outlined a simple form of miscue analysis by categorising the errors for substitutions, non-responses, insertions, omissions, pauses and correc tions, and then asking questions about the responses. Is the meaning still acceptable? How close is the miscue to the sound and shape of the printed word? Is the substi tution the same part of speech?
Southgate et al (1981) used miscue analysis to analyse the reading strategies of seven to nine year olds. Arnold (1982) suggested "a modified method of analysing miscues which should be feasible for general use." She provided a system of miscue analysis that could be used in the classroom and that had been discussed with several groups of teachers. Categories of miscues were analysed for non-response, refusal, omission, insertion, reversal, self-correction, hesitation and repetition. Substi tutions were grouped for their graphophonic, syntactic and semantic similarity to the words in the text. A Venn diagram also recorded whether each child might be using the cueing system negatively or positively.
Many of the reading schemes published during the 1970s and 1980s have focussed on language as it is used by children. Such language has been incorporated into the
text of these reading schemes. References to the ideas behind miscue analysis are to be found in some of the teaching manuals of reading schemes, for example Reading 360 published by Ginn. The manual states that "the limited use of graphic/sound cues produces successful reading when it supports the use of grammatical and meaning cues."
Bilingual children, however, may not have the facility with English to be able to take advantage of the approach. The content of the scheme may also be such that the children may not readily be able to relate to it. Some differentials in the miscue variables of monolingual and bilingual children might therefore be expected.
2.2.3 The criteria for the identification of the miscue variables
The variables are summarised in the table below. The criteria for the identification of each variable is afterwards described.
TABLE 1