Spivey, N. N., & King, J. R. (1989).
Readers as Writers Composing from Sources. Reading
Research Quarterly, 24 (1), 7-26.
Introduction
In discourse synthesis, readers/writers select, organize, and connect
content from source texts as they compose their own text. This study's purpose
was to examine the differences in students' performance on a report-writing
task that required students to gather information from source texts.
The study examines how the developmental patterns differed across three
grade levels and how students of different reading abilities at those grade
levels make use of sources to write their own informational reports.
The study focused on the textual evidence of selecting, organizing,
connecting, and the overall quality of the text.
Three questions guided the study: (1) On what basis do readers/writers
make selections for their synthesis?, (2) How do readers/writers organize the
content they select?, and (3) How successful are readers/writers at transforming
the source content into their own connected discourse?
Method
This study involved 60 students in sixth, eighth, and tenth grade
English/Language Arts classes in a public school in Texas. Each grade level
consisted of twenty students: ten were proficient readers for their grade level
who scored in the upper three stanines of the Reading Comprehension Subtest of
the Comprehensive Test of Basic Skills. (CTB) and ten were less proficient
readers for their grade level who scored in the lower three stanines of the CTB.
Over a three-day period, students wrote informational reports about the rodeo. Before writing, the students were to activate prior knowledge by brainstorming everything they knew about rodeos, then they were given three context sources selections Britannica Junior Encyclopedia, World Book Encyclopedia, and The New Book of Knowledge. All content sources were organized with collection structures of information clusters on the themes of rodeo events, rodeo sponsors, and participants' skills. During writing, the students developed their content for their written report from the encyclopedia sources; however, the written report need to be in their own words.
The written reports were measured on four dependent variables: (1) quantity of content which were the number of content units presented in the final written report, (2) organization which was measured with a breadth/depth ratio of numbers of thematic chunks to number of content units, (3) connectivity which was measured with a ratio between the number of connective inferences such as adding a referent or resolving an ambiguity that might be required of a reader and the number of content units, (4) holistic quality or overall quality of the whole report and three measures of task management: (1) planning which was measured by the written plan and markings, (2) retranscription which was measured by the changes within and across drafts, and (3) time which was measured by the number of minutes students spent on the task over the three-day period.
The written reports were analyzed by using three individual variances (ANOVA) which compared the students' selections with the textual sources. The first ANOVA was to determine whether students differed by grade level or reading ability in selecting intertextually important content. The second ANOVA was to compare the students by grade level and reading abilities on their reliance on source content in their writing. The third ANOVA was to test for differences by grade level and reading ability in the degree of reliance on a single textual source. In addition, the written reports were analyzed using two separate multivariate analysis of variances (MANOVA) for the differences on text measures and task measures.
Results
The analysis of the research data regarding the results of the first ANOVA
revealed a significant effect for grade level and reading ability with no
significant effect for the interaction between grade level and reading ability.
The older students included more content from the sources compared to the
younger students and the proficient readers included more source content
compared to the less proficient readers. The
older students and proficient readers were more inclined to include the content
that was most important intertextually.
The results of the second ANOVA revealed a significant effect for reading ability but not for grade level with a significant interaction effect between the reading ability and grade level. Whereas, the less proficient readers' reports had greater proportions of content added by them compared to the proficient readers' reports having greater proportions of direct source content.
The analysis of the third ANOVA presented no significant effect for reading ability or grade level nor was there any significant interaction effect. The source most used for content information was the World Book, which contained less original content than The New Book of Knowledge, but more than Britannica Junior.
The results of the analysis of text measures using the MANOVA revealed significant effects for both grade level and reading ability with no significant interaction effect. The sixth grade students differed from the tenth grade students in quantity of content and holistic quality, and differed in connectivity from both the eighth grade and the tenth grade students, who did not differ from each other. The tenth grade and eighth grade students' reports were more elaborate in content, were better connected, and were rated higher for holistic quality compared to the sixth grade students. There was a significant difference between the proficient readers and less proficient readers on all four measures that is quality of content, organization, connectivity, and holistic quality. The proficient readers' reports were more elaborate, more organized, better connected, and rated higher for holistic quality compared to the less proficient readers.
The results of the analysis of task measures using the MANOVA revealed a significant effect for reading ability but not for grade level with no significant interaction effect. There was a significant difference between the proficient and less proficient readers regarding planning and time. The proficient readers produced more elaborate written plans and spent more time on the task compared to the less proficient readers.
Discussion
The study found developmental differences in students' selections of
content from sources. The developmental patterns of selecting and connecting
content were evident in the grade levels. The
tenth graders as well as the eighth graders produced text with more content and
included important intertextual information, that is pertinent information that
was common across all sources, compared to the sixth graders.
The tenth and eighth graders written reports had strong coherent
representation in the text compared to the sixth graders, which presented many
gaps.
The study found performance differences in the proficient readers compared to the less proficient readers. The proficient reader's reports were rich in source content and include more source information from all three context sources selections compared to the less proficient reader. The proficient reader generated text with more connections and organization compared to the less proficient reader. The proficient readers developed a strong ability to organize information to develop a new text instead of reduplicating the source text compared to the less proficient reader. Also, the proficient readers were making more elaborate plans and expending more time when developing their reports.
In conclusion, the study found differences according to reading ability and grade level in the overall quality of the reports. The general reading ability and achievement in synthesis overlapped to a greater degree. When composing from sources, the student becomes the reader/writer, who is involved in the reading and writing process that is interconnected making it difficult to distinguish what is being done for the purpose of either reading or writing.
[Return to Additional References]