Many studies have reported associations between childrens' metalinguistic abilities and, in particular, their phonemic awareness capacities obtained prior to reading instruction, and their later level of reading achievement. Phonemic awareness is the child's ability to consider, reflect and generally manipulate the structural features of spoken language. This study investigated whether gender differences existed in the phonemic awareness capacities of preschoolers prior to commencing any formal program of reading instruction in an attempt to understand why young girls are generally better readers than young boys.

A group of 34 preschoolers with a mean age of 55.2 months and comprising 18 girls and 16 boys completed two phonemic awareness tasks focusing on the area of segmentation. A word segmentation task involved the children isolating and articulating the initial phoneme of a word. A second task called segment identity determined childrens' awareness of the identity of initial phonemes in different words. Information on childrens' letter and sound knowledge, a general index of their language development and an indicator of their performance on a control task of language ability were also obtained. The independent variable was gender while the dependent variables were performance on the two phonemic awareness tasks and the control task of language ability.

A range of inferential statistical procedures generally revealed no significant differences in phonemic awareness between boys and girls. No significant gender differences also emerged on the control task of language ability. When major subject background variables such as age, level of language development, letter and sound knowledge were controlled for conjointly, there were no significant differences between boys and girls in performance on either segmentation task.

The results are discussed in terms of young childrens' phonemic awareness capacities and the implications of the findings for understanding gender differences in reading acquisition are considered. Significant gender differences may exist in other areas of phonemic awareness, for example, blending. Alternatively, preliterate boys and girls may commence school with similar phonemic awareness capacities. As they progress through the early childhood years, however, boys may perceive reading to be more appropriate to girls through a combination of general social and cultural factors operating together with specific features of schooling, including classroom experiences or particular components of curriculum programs.

As the review of literature (Chapter Two) shows, a considerable amount of chronometric research has been undertaken on the addition facts over the last fifteen years. In many of these studies researchers fitted regression models involving structural variables to children's average latencies on the addition facts and interpreted different models as showing automaticity of addition facts or methods of calculation of answers such as counting on. From these equations some researchers have extrapolated trends across grades in the development of automaticity of recall. The review shows that a debate has arisen between researchers as to whether it can be assumed that children have automatised the answers to the addition facts or ways of calculating these answers. It shows that there is substantial evidence that children (and possibly adults) use a variety of strategies or procedures other than recall to derive answers which may persist over many years of schooling. The review also examines research on whether the use of such strategies can be expected to lead to automaticity of recall of answers to the addition facts, and also, on the question of how non-recall strategies are developed by children.

The present study examines the performance of 130 children from Grades 2 through 6 in one metropolitan primary school on the 100 addition facts and also on more complex addition problems. Children completed all tasks on a microcomputer with sums presented in a vertical format, and typed in their own answers to sums on a numeric keypad. The methodology and the types of sums are detailed in Chapter Three of the study. Latencies for every keystroke, the solutions given and correctness of answer were recorded on the computer. To relate latency data to the strategies used by each child, verbal protocol data of solution strategies were also obtained from each child on a number of addition facts. These were obtained after the computer task for the addition facts and were not intended to provide retrospective descriptions of the strategies used during the computer task, but to ascertain the range and nature of strategies which a child was using. The results of the verbal protocol analyses are reported in Chapter Four. Perhaps the most interesting finding of the verbal protocol data was that only three of the 130 children had automatic recall of answers for all 100 addition facts, one child from each of Grades 4, S and 6. It would appear that their automaticity was due to influences extraneous to the classroom. At this point it is important to note that the State primary mathematics syllabus which the school followed, recommended that the teaching of non-recall strategies for the addition facts in the early grades, based on an assumption that automaticity would developed from this by the end of Grade 5. Not only did this not occur, but examination of the verbal protocol data suggested that the non-recall strategies used by children were developed by themselves and did not apparently occur as a result of teaching.

Chapter Five briefly discusses the errors made by children on the addition facts and shows that the percentage of errors was small, and, except for the Grade 2 children, reasonably consistent across grade. Most errors occurred on facts with at least one addend of 7, 8 or 9, and sums greater than 11.

Chapter Six presents the main analyses of the study, the relating of latencies on addition facts to solution strategies. Since so few children had full automaticity of recall of addition fact answers, the question arose as to how the relative cognitive efficiency of the performance of children could be established. Since it was known that a variety of strategies were used by children in obtaining their answers, averaging of latencies and fitting of structural models was an inadequate means of analysing performance. (Following the procedures used by previous researchers one structural equation, the 'min' model, was calculated for each grade, however. This served two purposes: the fitted equations showed that overall the children in this study performed similarly to the children in previous research studies; and that the use of a production format yielded valid and reliable data.) At the time these analyses were being undertaken, Siegler (1987a, 1987b) warned of the perils of averaging latencies produced by different strategies. However, Siegler's approach of fitting different structural models to facts grouped by known solution strategy was not possible since retrospective verbal protocols were not obtained for the 100 addition facts. Instead, the addition facts were grouped, on the basis of previous research and the verbal protocol data of this study, into clusters for which particular solution strategies might be used. For example, in addition to the zeros, ones and doubles clusters which previous research had shown not to fit the main structural models, new clusters included near doubles and 9+ facts. Eight discrete clusters were used. By looking at children's average latencies on each of these clusters, a graphical 'processing print' was obtained for each child. Detailed case studies for twenty children showed that these prints were tied to the solution strategies nominated by children in their verbal protocols. On the basis of the processing prints and associated nominated strategies, six levels of cognitive efficiency of performance were designated for the children in this study with the most efficient level (children predominantly using automaticity of recall) being level 1, and the least efficient level (predominant use of very slow counting on or counting all strategies), level 6. While there was a relationship between grade and efficiency level, there was also considerable overlap, for example children from Grade 6 were classified at levels 1, 2, 3 and 4.

The last results reported in Chapter six explore the relationship between efficiency of processing of the addition facts, according to efficiency levels, and performance on the complex addition tasks. The analyses confirmed that inefficient processing of the addition facts was associated with inaccurate performance on these complex tasks. Analyses of the errors made showed that most errors were addition errors rather than inaccurate use of the addition algorithm. Perhaps the most interesting finding in Chapter Six is for performance on sums involving up to three columns and three digits which only Grades 5 and 6 children completed. Analyses by efficiency level showed significant differences in accuracy on the complex sums, errors which were predominantly addition errors rather than algorithmic errors, whereas analyses of performance on the complex sums by grade showed comparable performance for Grades 5 and 6. Therefore efficiency levels successfully identified clusters of children from these two grades who processed differently.

The final section of the study presents the results of an intervention study undertaken with two Grade 3 classes, one of which took part in the original data collection. The design of the study is detailed in Chapter Seven. The purpose of the intervention was to improve the automaticity of recall of addition fact answers for an experimental group drawn from the two classes and to see if this resulted in improved performance on more complex sums. A game format was used to present and practise addition facts in small groups over four weeks. The second computer data collection took place after the completion of the activities.

Chapter Eight reports the rather disappointing results of the intervention study. There was no significant change in performance on either the addition facts or the complex sums due to the intervention activities. However, a positive outcome of the intervention study was the validation of the processing print approach to determining processing style and efficiency. As graphs in Chapter Eight demonstrate, many children were remarkably consistent in their solution approaches over the three month period.

The findings of the study are summarised in Chapter Nine which raises many more questions for future research. It is believed that the study sheds new light on how children develop non-recall strategies for solution of the addition facts and the role of such strategies in the attainment of automaticity of recall of addition fact answers. The study establishes that there is a relationship between efficient processing of addition facts and performance on complex sums, where inefficient children appeared to handle the conceptual algorithmic requirements of the task (such as 'carrying') without difficulty but did not have sufficient capacity to process the minor component of the task, the mental addition of two or more digits. Finally, it is believed that the procedure of defining efficiency level could be incorporated into a computer-based diagnostic program which could give information not only on a child's accuracy, but also an indication of the efficiency with which they are deriving solutions to the addition facts.

The first aim was addressed by surveying the literature on methods of evaluating student performance in English. These approaches were discussed within a framework of three categories developed specifically for this study.

For the second aim, it was necessary first of all to identify and evaluate additional standards specifications developed since the implementation of criteria-based assessment in Queensland. A search of the literature revealed only one such theoretical model, namely McMeniman's (1986) Standards Schema. It is argued that her approach to standards is inappropriate to standards specifications for Senior English, and that an alternative was necessary. Developing and trialling this alternative formed a major part of this study.

The standards package which was compiled recognised the complementary nature of standards conveyed through verbal descriptions and exemplars (Salder, 1987). Compilation involved teachers of Senior English at All Hallows', a girls' Catholic school in Brisbane, together with the Year 12 student population. Teachers were involved in discussions which addressed the numerous issues in the use of exemplars. The students participated in compiling the folios from which were selected the exemplars used in the standards package. Each student completed a written questionnaire, and a representative cross-section were interviewed. The findings of both surveys and interviews were incorporated in a discussion of the standards package from a student perspective.

Multiple copies of the package were also distributed to 35 subject masters and teachers from a wide range of State and independent schools both within and outside the Brisbane region. The distribution was followed up by a seminar, the specific aims of which were to bring together all the recipients of the package to explain the approach taken and to provide an opportunity for the materials to be discussed and evaluated. Participants provided oral and written comment indicating their views on the usefulness of this form of standards specifications.

Responses to the standards package indicated a high level of teacher interest in and acceptance at both the materials trialled and the approach taken to standards specifications. An important feature of this descriptive approach was that it contextualised the exemplars by using the appropriate school work program and descriptive reports as support materials. The descriptive reports which established an important link between the exemplars and the requirements for the various levels of achievement as stated in the syllabus received favourable comment. The trial also revealed a high level of consensus among participants of the sampled group that the exemplars illustrated accurately the syllabus achievement levels. On the basis of this it was concluded that the communication of standards in English by way of exemplars and verbal descriptions represents a workable model for standards specifications.

The implications of the results of the extended theoretical analysis into standards specifications in English in the first section of the study, the empirical testing of some of the theoretical arguments in the second section of the study, and the advantages of communicating standards to students are discussed in turn. It is recognised that any approach to standards must consider not only the appropriateness of the method of indicating standards, but also the suitability and effectiveness of various means for promulgating the particular method chosen. The latter, which is considered central to the long-term success of any standards model, represents an important area for future research.

The 17 intending teachers who participated in the study formed the data source through which the following research questions were explored:

(1)	Over the period of pre-service teacher education, how do intending teachers construe their experiences of becoming teachers?
(2)	Does change occur in how individuals construe the experience of becoming a teacher over this time?
(3)	Where change does occur:
	(a)	What is the nature of that change both for the research group as a whole and for the individuals who make up the group?
	(b)	Does this change correspond to the form of development previously identified in the literature on preservice education?
	(c)	Can change be promoted through the use of 'fixed role treatments'?

To answer these questions, a variety of methods was used, largely based on personal construct theory (Kelly, 1955). In employing these diverse methods, the overall emphasis was on the exploration and elaboration of individuals' perspectives on their experiences of becoming teachers and how these perspectives changed over time. The repertory grid employed in the study was designed so that participants construed the experiences associated with pre-service teacher education. The grid was elicited from the 17 intending teachers on three occasions, following their periods of practice teaching in schools. The grids served as a repeated measure on which to structure the sequence of report and reflection that all 17 participants engaged in with the researcher across the entire preservice period.

In addition, other data collection methods were selected that complemented this emphasis on the individual and change. Following each of the three grid elicitations, repertory grid data were analysed and presented to the intending teachers in individual 'learning conversations' with the researcher.

Additional methods of enquiry and data collection included that of a writing program, in which the participants wrote in a variety of forms on the experience of becoming a teacher. Four of the 17 participants also took part in fixed role treatments. In this intervention, individuals whose repertory grids and learning conversations suggested a decline over the year in their perception of self-as-teacher enacted role scripts at their practice schools that depicted them as successful teachers. The role scripts were based on that person's perception of successful teachers, as derived from the individual's repertory grids, learning conversations and writing program. Additional data in the form of observational records of campus and practice teaching performance were available to the researcher, who was also one of their teacher education lecturers and therefore in daily contact with the 17 intending teachers.

Data from this range of sources enabled answers to be presented to research questions (1), (2) and (3a) on individual change, and to research question (3c) on fixed role treatments. In regard to questions (1), (2) and (3a), the data relating to the individual intending teachers demonstrated that change did occur over the period of their pre-service program. This change was shown to be unique to each participant, by reference to the whole range of data sources. Fixed role treatments, the topic of question (3c), were shown by both qualitative and quantitative means to have assisted the targeted participants to perceive themselves as more successful teachers.

By using the SOCIOGRIDS computer program, similarities in construing among the individual repertory grids were identified, so that a mode grid could be produced for each of the three elicitation occasions. Data from the learning conversations, the writing program and other sources explicated the constructs that were supplied by those individuals whom the SOCIOGRIDS program identified as being mode construers on each of these occasions. A group of experienced teacher educators matched these explicated constructs against Fuller's stages of teacher concerns. The teacher educators' assessment indicated that, for this group of intending teachers, change had occurred in the form that Fuller had specified. This provided an answer to research questions (3a) and (3b) on group change.

In finding answers to these questions, a number of other research assumptions have been reinforced. Firstly, the data provided by the study support the belief that individuals make idiosyncratic meanings of what is offered in teacher education courses and that these are not open to outsider prediction. A second conclusion is that the very unpredictability of personal meanings makes it imperative that an assortment of methods be used to ascertain these constructs. Finally, if a study aspires to represent the meanings that individuals make of becoming a teacher, participants must be continually involved in the research activity, so that they have the opportunity to verify their interpretations of their experiences. Becoming a teacher is an important phase in an individual's professional life and as such this period should be recorded with integrity and authenticity.

Please cite as: QIER (1989). Thesis abstracts. Queensland Researcher, 5(3), 41-53. http://www.iier.org.au/qjer/qr5/thesis-abs-5-3.html