A study to determine degree of inferred story-character identification and its resultant effect in reading comprehension in selected multiethnic reading texts

No Thumbnail Available
Authors
Messmore, Peter B. (Peter Burl), 1940-
Advisor
Sciara, Frank J.
Issue Date
1971
Keyword
Degree
Thesis (D. Ed.)
Department
Other Identifiers
Abstract

This study attempted to investigate the relationship between inferred identification and two types of reading comprehension, molar and specific, at the third grade level. Subjects for this study were selected from six schools located in four midwestern urban centers. The 121 Negro and Caucasian children who participated in the study attended schools designated as Title I schools.Pre-tests were administered in the Fall of 1970 to determine intellectual level and pre-molar reading comprehension. Three administrations of the Modified Semantic Differential for Elementary Children were given to determine identification level as well as three administrations of the Specific Story Comprehension Test to assess specific story comprehension. Post-test administration of molar reading comprehension test was conducted in February, 1971. Reading materials consisted of three stories selected from multi-ethnic texts which were in use in each of the schools.Two covariates, pre-molar reading comprehension scores and scores from the Lorge-Thorndike Intelligence Test, and two dependent variables, post-molar reading comprehension scores and averaged specific reading comprehension scores, served as a basis for statistical analysis. Testing of nine hypotheses resulted in data which formed the basis for two analysis of variance tables for each dependent variable; F values were obtained for the following sources of variation: (1) race, (2) sex, (3) identification, (4) race/sex, (5) race/identification, (6) sex/identification, and (7) race/sex/identification. The dependent variable, molar reading comprehension, produced only one F value which was significant at the .05 level of significance; this value was found for the source-race/identification. A multiple comparison technique was applied to determine which race and which level of identification were significantly related to molar comprehension and results suggested that both Negro and Caucasian children operating at the high and moderate identification levels were doing so at significant levels (eleven comparisons were significant at the .01 level of significance; two comparisons were significant at the .05 level of significance). Comparisons which involved Negro and Caucasian children identifying at the moderate level with Negro and Caucasian children identifying at a low level were not significant. Analysis of data from the second dependent variable, specific reading comprehension, failed to produce any significant F values.The finding in this study that children who identified at high and moderate identification levels achieved significantly higher molar reading comprehension scores than low-identifying children of both races seemed to offer some justification for recent efforts to include diverse ethnic characters in elementary reading texts. A more thoughtful examination of this relationship however, may indicate the possible presence of a third and as yet unknown factor which acted in conjunction with the identification process and to which the gain witnessed in molar reading comprehension may be attributed. The original impetus for this study, the implied suggestion of several authorities that disadvantaged children need specific stories with which they can identify before effective reading can occur, appeared to be answered by data found on the second dependent variable, specific reading comprehension. The finding that there were no significant differences for parts of the nine hypotheses relating to specific reading comprehension would seem to suggest that inclusion of diverse ethnic story-characters in elementary reading texts because disadvantaged children can more easily identify with them and subsequently aid their reading comprehension may not be justified by the analysis of data obtained in this study.These results may indicate that the well-meaning attempts to foster reading comprehension of disadvantaged American children during the mid-1960's by including diverse ethnic characters in materials with which they learn to read may be a somewhat simplistic solution to a highly complex problem. Further, it appeared that identification may be only one component of what might be called an affective-gestalt and that application of these affective components into elementary reading texts required greater sophistication and research insights than we now possess.