Cognition, Metacognition, and Achievement of College Students with Learning Disabilities

This study examined the way successful college students with LD compensated for their deficits in phonological processing. Successful was defined as average or above-average grades in college coursework. The study compared the cognitive and metacognitive performance of students with and without LD (N=40). Although achievement levels for both groups were comparahle, students with LD scored significantly lower than students without LD in word reading, processing speed, semantic processing, and short-term memory. Differences were also found between groups in self-regulation and number of hours of studying. Results showed that students with LD compensated for their processing deficits by relying on verbal abilities, learning strategies, and help seeking. Guy Trainin, Ph.D., Univer5ity of Nebraska-Lincoln. H. Lee Swanson, Ph.D., University of California-Riverside. Despite their persisting academic difficulties, adults with learning disabilities (LD) are enrolling in postsecondary education in increasing numbers (Newman & Cameto 1993; Henderson, 1999; Wagner & Blackorby, 1996). A major impetus behind this growth has been legislation and educational reform in K-12 settings. These changes in public policy, in turn, have increased funding, services, and efforts to identify and serve students with LD in school, thus producing more high-school graduates who are eligible for college entrance. Parallel to these developments, postsecondary institutions now grapple with the issues of serving college students with LD, guided by section 504 of the Rehabilitation Act (1977) (as quoted in Adelman & Vogel, 1993) and the Americans with Disabilities Act (ADA) of 1990. As a result of the increased efficacy of K-12 education for students with LD, this group is now the largest group of college students with disabilities (Henderson, 1999). Research suggests that college students with LD have significant difficulties in multiple academic domains (e.g., Vogel & Adelman, 1992; Wilczenski, 1993). Forexample, reports on students with LD indicate problems in reading, writing, math, and foreign language. They are at risk for failure in their courses and at increased risk for dropping out of college beyond the freshman year compared with their nonlearning disabled (NLD) peers (Vogel & Adelman, 1992; Wilczenski, 1993). At the same time, however, studies on colleges with adequate academic support show that students with LD can attain normative levels of achievement (McGuire, Hall, & Litt, 1991; Vogel & Adelman, 1992). The question is how do students with these difficulties compensate for their deficits? One of the ways college students with LD may compensate for their cognitive difficulties is by relying on metacognition; that is, consciously controlling actions Volume 28, Fall 2005 261 that are too complex to be controlled automatically. Metacognition is defined by Zimmerman (1986) as the ability to adjust behavioral, environmental functioning in response to changing academic demands. Pintrich (1994) defines academic metacognition as a construct comprised of three major elements: (a) active control over learning-related behaviors such as when, how much, and with whom a student is learning; (b) self-regulation of motivation and affect, in which students learn how to control their emotions and even use them in goal setting; (c) control over various cognitive strategies for learning, such as rehearsal and memory strategies. Use of metacognitive strategies may be linked to efficient ways to improve performance in academic and work environments. For example, young adults in college regulate their learning behavior with a variety of specific strategies (Pintrich, 1994). They manage their own time, decide with whom to study, and monitor their comprehension with a variety of internal (selfregulation, strategies) and external (peers, family, faculty) supports. In addition, several studies have shown that metacognition is not a set of idiosyncratic behaviors but a finite set of common skills that are highly correlated to academic success (Garcia & Pintrich, 1994; Pintrich, 1994). Reis, McGuire, and New (2000) described qualitative patterns focusing mainly on motivational aspects in support of the hypothesis that metacognition plays a major role in the success of students with LD. Recent research on college students with LD suggests that metacognition (e.g., self-regulated behavior) is a strong predictor of academic success (Ruban, 2000; Smitely, 2001). In explaining the compensatory role of metacognition in the academic achievement of students with LD, an analogy may be drawn to research on the relationship between metacognition and cognitive deficits in the literature on aging. While older adults face diminishing cognitive abilities and reduction in efficiency (Salthouse, 1992), some fare much better than others for a longer time (Dixon & Baeckman, 1995). Baltes and Baltes (1990) suggested three possible paths to compensating for lost cognitive abilities. One such path is metacognitive control, defined as strategies, effort, and self-regulation of activities. This path must be under the direct control of the individual. That is, the person faced with a task related to a deficit must decide to use a strategy to compensate while weighing costs and benefits. This includes strategies related to the individual's actions (e.g., making lists to compensate for memory deficit), as well as strategies that include technological supports and help seeking. Another path, cognitive compensation, implies that individuals who face a deficit in one cognitive domain may use relative strengths from another cognitive domain such as general processing, working memory, or semantic ability that can assist performance. Finally, Dixon and Baeckman (1995) offer a third option, lowering of standards. Here, the individual chooses an alternative approach that reduces the demands on the deficient process; for example, writing shorter sentences to reduce working memory load (Kemper, 1986). In this study we considered whether successful students with LD rely on metacognitive strategies to compensate for processing difficulties. We also considered the deficient cognitive processes that metacognition may compensate for. The literature suggests that students with LD in most cases have deficits in phonological processing. For example, comparing the phonological processing of adult LD with age-and-readinglevel matched groups, Bruck (1992) found that adults with childhood diagnoses of LD functioned far below the norm for both their age and reading level. Similarly, Scarborough (1984) reported that adults with a childhood history of dyslexia performed poorly on non-word spelling tasks, confirming the persistence of phonological encoding difficulties. A study by Pennington and his colleagues (1986) also revealed a deficit in phonological processing in adults with LD, while showing normative performance in their ability to spell morphologically complex patterns. In a subsequent study, Lefly and Pennington (1991) found that the same pattern of low phonology in adults with LD was compensated for by spelling. Although not ageappropriate, spelling of morphologically complex words was better than phonological spellings. Considerable evidence shows that the word reading ability of adults with LD is consistently lower than that of their nondisabled peers (Apthorp, 1995; Bruck, 1990; Ritter, 2000). While the reading comprehension of college students with LD was considerably lower than that of their NLD peers, it was still considerably higher than that of decoding level-matched (hence younger) participants (Bruck, 1985, 1990; Ransby & Swanson, 2003). Furthermore, comprehension of both reading and listening has been shown to be orthogonal to reading fluency and word reading ability in LD students (Ransby & Swanson, 2003). This finding suggests that college students with LD may rely on different cognitive processes (e.g., processes related to listening comprehension rather than phonological decoding) in their path to reading comprehension compared to their NLD peers. In summary, the purpose of this study was to explore whether college students identified with LD use metacognitive strategies to compensate for their phonological processing deficits. Two student groups (LD vs. NLD) Learning Disabiiity Qiiarteriy 262 were compared to determine those processes that are similar and those that are different between the two groups. We hypothesized that college students with disabilities compensate in cognitive processing by relying on metacognitive strategies. We also explored whether such students rely on strengths in other cognitive processes in addition to metacognition that allow them to compensate for difficulties in phonological processing. METHOD Participants Participants were 40 students (20 LD, 20 without LD) from four universities in Southern California. Students with LD volunteered for the study after a letter soliciting their participation was sent by the disability centers in the respective universities. Students without LD (referred to as NLD) were matched as closely as possible to the population of students with LD on demographic variables of ethnicity, college major, gender, and academic standing (i.e., grades). Participants' chronological age ranged from 18 to 64 years. The NLD group had a mean age of 21.75 (5D=1.12), while the group with LD had a mean age of 31.4 (5D=13.56). The LD group consisted of 15 women and 5 men. The control group (NLD) consisted of 16 women and 4 men. Overall ethnic composition was 40% White nonHispanic, 20% Hispanic, 12.5% African American, 22.5% Asian American, and 5% other or mixed ethnicities. The distribution of ethnicity in the two groups was not significantly different, x^(4)=3.95, p=.41. Class standing in the LD and NLD groups was similar, as illustrated in Table 1. Further, there was no difference between the distributions of major fields of study between the groups. Participants' socioeconomic status was measured by parental education. The NLD group had a higher level of parental education, M=5.53 (5D=1.29) (5=some postsecondary education and 6=undergraduate degree), compared to students with LD (M=4.22 (SD=2.15). The difference was statistically significant t{38)=2.32, p-.O27, d=.73. Most of the difference between the LD and NLD groups may be attributed to four participants (20%) in the LD group who came from immigrant families and whose parents had either an elementary education or no formal education at all. The observed difference corresponds to the data on young adults with disabilities presented by Wagner and her colleagues (1993). Selection Students with LD were recruited with cooperation from the university disability centers at participating Table 1 Class Standing and Major Field of Study for LD and