Composition revision for grade four pupils
Date
1976
Authors
Mitchell, Phyllis May Jeanne
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
The purpose of the study was to compare the effect of three treatments on the writing of boys and girls in Grade Four. The enthusiasm and willing participation of District teachers reflects a growing concern regarding the effectiveness of current methods of writing instruction.
A random sampling of two boys and two girls was taken from each of the twelve classes involved in the study. The total sample of 48 pupils consisted of 24 boys and 24 girls. An examination of the Blishen socio-economic index scores computed for the sampling reveals a mean score of 49.
Twelve Grade Four teachers employed in ten Greater Victoria schools volunteered to participate in this study. Their classes were randomly assigned to one of three experimental groups. As described in the manual prepared by the researcher, the teachers were required to present ten writing experiences to their classes during the ten weeks of the treatment.
During each week of the treatment two lessons were presented. The format of the lessons was such that the time allotted for presentation by the teachers and writing by the pupils was equivalent for all three groups. In the first lesson in each week, the stimulus developed in the teachers' presentation was identical for all three experimental groups, but in the second lesson, during the teachers' presentation, the procedure for evaluation and the method of skills instruction was different for each experimental group.
At the beginning of the study in September, two composition pre-tests (one descriptive and one narrative) were administered by the teachers. After completion of the treatment the teachers administered two post-tests (one descriptive and one narrative). The spelling on these compositions was corrected by the researcher and the tests were coded for the benefit of the researcher only. The tests were then typed to preserve the anonymity of the pupils.
Three raters were employed by the researcher to evaluate the composition tests. At a training session, the raters practiced their evaluation of the composition tests according to the criteria they had determined. An overall correlation coefficient of 77 among judgements (significant at the .01 level) was determined.
Correlated "t" tests and "F" ratios we-re used to summarise the findings and to test for significance. Correlated "t" tests were used to measure pupils gains in quantity and quality of writing. There were no significant differences among treatment groups in pupils gains. The effect of treatment and sex on ·pupils' gains in quantity and quality of writing was investigated using two-way analysis of covariance. "F" ratios were used to determine the significance of the effect of these two variables. No significant differences were found.
Some interesting directions were noted by the researcher in a closer analysis of the data.
In respect to quality of writing, the girls achieved consistently higher scores than the boys at the beginning of the study. This suggested the need for further research to determine the developmental characteristics of boys and girls in written language achievement.
A series of stimulus lessons was developed by the researcher to motivate pupils to write. There is some evidence in this study to suggest that boys and girls had different composition interests. The boys in each treatment group made particularly significant gains in the quality of narrative writing whereas the girls in two treatment groups made significant gains in the quantity of descriptive writing. Further research to determine the intensity and variety of composition interests of boys and girls was recommended.
The results of this study would seem to invite further research into developmental patterns among boys and girls in building writing competence. In addition research into the intensity and variety of composition interests among boys and girls should be considered. Finally, future studies using larger samples and longer terms may well provide more conclusive findings.