index by author
index by subject
| |
home index by author index by subject |
Friday stream 6 Session 11.35 - 12.50 Length 25 minutes
Beverly A. Warde
Department of Exceptional Student Education, Florida Atlantic University, U.S.A bevwarde@yahoo.com
Abstract
This study examined oral and silent reading comprehension and oral reading miscues in an historical text structure. Forty college students with and without learning disabilities in reading were asked to read silently and orally two historical text structure passages from a college level textbook. Oral reading miscues and oral retellings were analyzed. Results revealed that the college students with learning disabilities in reading produced a significantly greater total number of oral reading miscues and earned significantly fewer comprehension points than their peers without learning disabilities. Results are discussed in light of their implications for curriculum and instructional planning for school-age students with learning disabilities, and the implications for post-secondary service provisions for college students with learning disabilities in reading.
Reading at any level is a complex task composed of not only the decoding of the printed words, but also, more importantly, the drawing of meaning from the text read. At the college level it is essential that students be able to rapidly and efficiently decode multi-syllable words, and comprehend of many pages of content texts including recognizing and utilizing various text structures. Moats and Lyon (1993) found that 80% of the children identified as have a learning disability had their primary difficulty in reading and language. These problems continue into adulthood (Bruck,1985). Vogel and Adelman (1990) found that college students with learning disabilities (LD) had significantly lower word attack abilities when compared to peers. They do not use age appropriate word recognition processes (Bruck, 1990), and they are slower at identifying words in isolation and context when compared to their peers (Ben-Dror, Pollatsek & Scarpati, 1991). College students with LD read at a much slower rate than their peers (Runyan,1991), and their recall of information read from narratives is at the same level as third graders. (Bacon & Carpenter, 1989). From these studies we can see that college students with LD perform at a lower level than their peers on tasks requiring reading from isolated word lists, standardized timed reading tests, and recalling information from narratives. However, we do not know how the college students with LD perform when reading in a college text. A task they must do on a daily basis in college.
This study was designed to explore that issue of reading college-level, content-area text, and to describe the reading characteristics of college students with learning disabilities compared to their college peers without learning disabilities. In examining these characteristics, the study had two aims: (a) to determine if the students with LD differ from their peers in the quantity of miscues made while reading orally; and (b) to determine if comprehension scores (determined by retellings) of text read either orally or silently differed between the two groups.
Method
Participants
The participants were 20 college students with learning disabilities (LD) in reading (11 men and 7 women, mean age =25, mean GPA=2.8), and a control group of 20 college students without learning disabilities (NLD) (12 men and 8 women, mean age =24, mean GPA=2.85). All subjects were ranked as either juniors or seniors, had completed at least one semester at the university, and had retaken the CLAST Reading section. The CLAST (College Level Academic Skills Test), a timed achievement test covering mathematics, English grammar, reading comprehension, and a written essay, is required of all college students enrolled in public universities in the state of Florida upon completion of the core curriculum (at the end of the second year of college). Students who do not achieve the required score on any section may be retaken that section with double the initial time allowed. All subjects in this study had retaken the Reading section with double time with resulting mean scores of 307 for the students with LD, and 308 for the NLD students.
The two groups were controlled for year in college (junior or senior), college, and CLAST Reading score. Since a student's chosen major may influence the type and amount of reading completed in college, equal numbers of LD and NLD subjects were drawn from each college represented in the study. Within each college, the range of LD subjects' CLAST Reading scores was determined. The NLD subjects were then recruited from students within a particular college who had a similar range of scores (range for LD subjects, 293-328; range for NLD subjects, 295-330).
The students with LD all received support services from the Office for Students with Disabilities at Florida Atlantic University. To qualify for services and accommodations, a student had to present testing reports completed within three years that diagnosed a learning disability and complete an extensive intake interview with a learning disabilities specialist. Only students with documented learning disabilities specifically in the area of reading, and who had a history reading problems and identification as LD since childhood were included in the sample. The students in the LD group met the criteria of average to above average intellectual ability on the WAIS-R and a discrepancy of 1.5 standard deviations or greater between measures of aptitude and achievement in reading. All subjects with LD reported continued problems with reading and the need to reread material several times for maximum comprehension.
The control group was recruited from a pool of possible subjects who had retaken the Reading section of the CLAST twice, whose primary language was English, and who met the criteria of junior/senior status, CLAST score within the selected range, and appropriate college. This process produced a pool of 41 potential NLD subjects who were contacted by mail. Twenty students consented to participate.
Instruments
Reading Selections - The materials for the two reading passages were developed from a college textbook used in introductory sociology courses , Social Problems (Eitzen & Zinn, 1994). Each passage was a complete subsection within a chapter, and of an historical (time order) text structure. One passage, "Feminist Movements in the United States" (476 words), was designated as the oral reading selection, and the other, "Differences Among Ethnic Groups"(436 words), was the silent reading selection. In an attempt to control for the effect of prior knowledge, potential subjects majoring in history were not selected. Historical passages were chosen since this text structure is considered to be easier to comprehend than enumeration, compare/contrast, or cause/effect structures.
Comprehension Scoring Form - A comprehension retelling scoring guide was developed using a three-step process developed by Vacca and Vacca (1989). Five faculty members from the sociology and psychology departments who were considered knowledgeable about introductory social-psychology issues independently selected phrases that they thought reflected the main ideas and important details in each reading selection. The resulting lists of phrases were categorized by myself and a second rater as either "major points", "main supporting points", or "details". The passages were then evaluated for historical text signal words. From this process a final comprehension scoring guide was developed for each reading selection.
Miscue Analysis Scoring Form - A scoring form for recording the oral reading miscues was developed following similar forms developed by Goodman, Watson, and Burke (1987).
Procedure
All subjects met individually with the investigator for one session of approximately 30 to 45 minutes. Subjects were informed that they would be asked to retell the readings immediately after completion of the oral or silent reading task, and that the session would be tape recorded. Prior to reading a selection the subjects were provided an overview of the chapter from which each selection was taken. Passages were presented in random order. After completion of the tasks, each subject was asked questions, the answer to which provided background and demographic information.
Scoring - All scoring was carried out by the investigator and three qualified raters who had no knowledge of the subjects classifications as either LD or not. The raters, an instructor in college reading courses, a speech/language pathologist, and a doctoral student in linguistics, had been trained to proficiency using the samples from four practice students. To develop materials for training sessions for the raters, four students (two with LD and two without LD) who would have fit the criteria for subject selection except for class standing were recruited. The transcriptions of their oral reading miscues and silent/oral reading retellings were used as the training materials
The investigator and one other rater independently recorded reading miscues on the Miscue Scoring Form while listening to the tape recording of each students' oral reading. Differences between the two raters were resolved by jointly listening to the recordings at varying speed and tone settings on the tape recorder. Each subject's oral reading miscues were transferred to the Miscue Analysis Form, and then analyzed and categorized by the investigator and another rater. Interrater reliability was established for syntactically acceptable miscues (.95), semantically acceptable miscues (.92), and meaning change miscues (.92). The number of miscues each subject made while reading orally was totaled. Miscues that were coded as syntactically or semantically unacceptable, or caused a textual meaning change were counted as "loss-of-meaning miscues". This category of miscue was converted to a percentage of the total number of miscues made.
To score the comprehension, each subject's oral retellings were transcribed verbatim, then scored by two raters using the Comprehension Retelling Scoring forms. Interrater reliability was established for scoring the two passages at .90 and .89.
Data Analysis
The differences in total number of miscues and mean percentage for loss-of- meaning miscues were analyzed with a factorial analysis of variance (ANOVA) with repeated measures with one between-group variable (LD versus NLD). The differences in comprehension retelling scores of LD and NLD subjects were analyzed with a repeated measures factorial analysis of variance using retelling raw scores as the dependent variable, group membership (LD versus NLD) as the between-subjects factor, and reading method (oral versus silent) as the within-group variable.
Results
Analyses Oral Reading Miscues
Results of the ANOVA for the total number (raw score) of miscues revealed a statistically significant main effect for group, F (1,38) = 56.33, p < .0001. The college students with LD produced a significantly greater number of total miscues (M=26.15, SD=14.15) compared to their peers without LD (M=3.60, SD=2.39) For miscues that caused a "loss-of-meaning", there was a significant main effect for group, F (1,38) = 13.71, p<.001. The students with LD not only had a greater total number of miscues, but they also had a significantly higher percentage of miscues that would result in a loss-of-meaning (M=58%, SD=.18) when compared to the students with LD (M=26%, SD=.32).
Analyses of Comprehension
Results of the factorial ANOVA for comprehension revealed a statistically significant main effect for group, F (1,38) = 68.80, p < .0001. The college students without LD scored significantly more points for comprehension retellings for either the orally (mean=9.35, SD=3.43) or silently (mean=8.15, SD=2.20) read passages than the students with LD (oral selection mean = 2.20, SD = 2.37; silent selection mean = 3.25, SD = 2.46). An analysis for method (oral and silent) found no main effect for method. However, there was a disordinal interaction between group and method. The LD subjects earned more comprehension points when silently reading an historical text than when they read orally. The NLD subjects earned more points for their orally read historical text than for the silently read historical text. A subsequent paired t-test indicated that the difference in comprehension between orally and silently read historical texts was not significant for either group.
Discussion
The main purpose of this study was to determine if differences existed in the reading abilities of college students with a history of learning disabilities in reading when compared to students without learning disabilities who might be considered "poor" college readers since they did not pass the Reading section of the CLAST on the first attempt. Two facets of reading ability in an historical text structure were explored: word recognition and comprehension. Since an individual's reading comprehension might be affected by the method by which the material was read, both oral and silent reading measures were taken utilizing passages of an historical text structure. The important finding of this study was that college students with a history of a learning disability in reading still exhibit significantly poorer reading skills than their peers.
The most notable pieces of evidence in the search to differentiate the LD from the non LD college was the total number of miscues produced, and the higher percentage of loss-of-meaning miscues. The LD subjects were misreading a significantly greater number of words. The literature abounds with many years of research attesting to the reading decoding problems of children with learning disabilities. Only recently have college students with LD been the focus of research. The research studies involving the college age population supported the results of this study regarding the word recognition abilities of these students (Ben-Dror, Pollatsek, & Scarpati, 1991; Bruck, 1990; Kitz, & Tarver,1989). Considering the amount of print exposure that most young adults in college experience, it seemed surprising that deficits in word recognition persisted into adulthood. For most college students, word recognition in contextual reading is an overlearned and automatic cognitive activity. The subjects without learning disabilities, though considered "poor" readers, produced very few miscues.
Another goal of this study was to research the reading comprehension abilities of college students in as naturalistic a manner as possible. They were asked to retell what they remembered about each article after completing the reading without any predetermined questions, as they must do in reality. Given these conditions, the NLD students scored significantly higher on comprehension recall than the students with LD whether they read orally or silently. The LD subjects comprehension scores were higher when they read silently, leading one to conclude that, as with younger children with LD, the college students with LD were apparently expending a great deal of effort to decode the words at the expense of comprehension.
Historical text structure was chosen since it is considered to be easier than some other expository text structures, and it produces readily recognizable signal words (e.g. "before the twentieth century"). Given the significant difference between the two groups on comprehension scores, the subjects' use of structure cues was investigated by reviewing the transcriptions of retellings. All of the NLD subjects utilized the text structure to aid in recall for both reading passages, whereas, only 20 % of the LD subjects did. Thus, the poor comprehension performance of the college students with LD may be attributed to excessive miscues that change meaning, a lack of awareness of meta-cognitive strategies concerning text structure, or the failure to apply those strategies.
References
Bacon, E.H., & Carpenter, D. (1989). Learning disabled and nondisabled college students' use of structure in recall of stories and text. Learning Disability Quarterly, 12, 108-118.
Ben-Dror, I., Pollatsek, A., & Scarpati, S. (1991, May). Word identification in isolation and in context by college dyslexic students. Brain and Language, 40, 471-490.
Bruck, M. (1985). The adult functioning of children with specific learning disabilities: A follow-up study. In I. Siegel (ed.). Advances in Applied Developmental Psychology. New Jersey: Ablex.
Goodman, Y.M., Watson, D.J., & Burke, C.L. (1987). Reading Miscue Inventory: Alternative Procedures. New York: R.C. Owen Publishers.
Kitz, W.R., & Tarver, S.G. (1989). Comparison of dyslexic and non-dyslexic adults on decoding and phonemic awareness tasks. In R.F. Bowler (ed.), Annals of Dyslexia (Vol. 39, pp. 196-205). Baltimore, MD: Orton Dyslexia Society.
Moats, L.C., & Lyon, G. R. (1993). Learning disabilities in the United States: Advocacy, science, and the future of the field. Journal of Learning Disabilities, 26(5), 282-294.
Runyan, M.K. (1991). The effect of extra time on reading comprehension scores for university students with and without learning disabilities. Journal of Learning Disabilities, 24(2), 104-108.
Vacca, R.T., & Vacca, J.L. (1989). Content Area Reading. Harper Collins. Vogel, S.A., & Adelman, P. (1990). Extrinsic and intrinsic factors in graduation and academic failure among LD college students. Annals of Dyslexia, 40, 119-137.
 |
 |