EFFECTS OF AGE AND SCHOOL GRADE ON THE CO HERENCE OF A NARRATION WRITTEN BY CHILDREN WITH LEARNING-PROBLEMS

This paper studies the degree of coherence achieved by short narrative s written by children exhibiting learning disabilities related to read ing and writing. It reports the coherence in relation to the age and t he school grade of the children. It investigates to what extent the re sults are able to throw light in the pertinence of the two main theore tical approaches traditionally used to understand reading/writing disa bilities in children: the ''deficit'' model and the ''similar sequence '' model. To that effect, a sample of 60 children previously diagnosed a shaving trouble with reading and writing was chosen in the Psychopa degogic Center of the Ministry of Education. The first twelve children of this sample were in 2nd grade of primary school, the second twelve in 3rd grade, and so on up to the 6th grade. A text was elicited from them on the basis of a short story which was previously read aloud to them. A control group was then selected, carefully matched to the stu dy group with respect to school grade, school type and sex. As far as possible, both groups were also matched as to age. An original analysi s was specifically designed to mark the level of coherence attained by each one of the 120 texts. Besides, the number of ''propositions'', o r rather story points, recovered from the stimulus text, and the numbe r of words used to recover them, were also counted, as is more usual i n this kind of study. The results obtained demonstrate that the contro l children achieved higher scores in all three variables. However, cor relation and regression analysis show that both age and school grade a re highly explanatory of the three linguistic outcomes in the control group, whereas the study group presents a more differentiated picture. Variation in age in age and school grade of the study group are less significantly correlated with, and determinant of word counts and stor y point scores, and there is no significant correlation or determinati on in the case of the level of coherence. On the other hand, since the differences between both groups of children are more evident in the c ase of age than in the case of school grade, the results of this study deem to support the ''similar sequence'' hypothesis. On the other han d, the study suggests that in reading, as disabled children grow older and advance in school, might certainly learn to use more words and to recover more story points when constructing their narratives, but the y would fail to produce more coherent texts. If such a hypothesis is v erified in a longitudinal study, then one would have to postulate insu perable, specifically linguistic disabilities for such children.