| This study examined the relationship of assignment topic, purpose and audience to writing quality and use of lexical cohesion in essays written by selected students in seventh, ninth, and eleventh grades, and college freshmen. The variables of more and less specific information on topic, purpose and audience were manipulated simultaneously to determine their separate and combined effects on writing quality and lexical cohesion.;Analysis of variance was used to determine the effects of grade level and more or less specific information about assignment writing. Essays were also examined on the basis of words and lexical ties per T-unit, total number and frequency of specific lexical ties per essay, and an index of lexical density.;In general, results indicate that essays produced from more specific information about purpose, "complementary" amounts of information about topic and audience, and "opposing" amounts of information about purpose and audience were rated higher for most levels. Eleventh-grade writers utilized rhetorical cues more than other groups, while college freshmen used them less. Because of maturation and/or training, college students were not as dependent upon information in assignments as were younger, less experienced writers.;Total number of lexical ties increased with grade until the college level where fewer, more sophisticated forms (e.g., synonyms and collocations) were used. Although lexical cohesion analysis yielded modest results, the value of this method is limited by the pervasiveness of cohesive ties and the inability of cohesion analysis to identify discourse-level unity.;Sixty-four writing samples, one high-rated essay and one low-rated essay from each of the eight assignment groups at four grade levels, were chosen for analysis. Four raters trained in holistic evaluation techniques scored essays in two separate sessions.;The development differences among ages and the role of audience raise questions for further research. |
