The number of ESL students entering colleges and universities is steadily increasing.  These students need to develop English language competence to succeed in the mainstream academic curriculum.  In an attempt to facilitate the mainstreaming of ESL students into this academic curriculum, ESL reading courses have focused on the introduction of academic content-based readings. 

      Studies have shown that the introduction of academic content into ESL reading courses eases the transition of students into academic content classes (Benesch, 1988; Brinton, Snow, & Wesche, 1989; CUNY Language Forum and CUNY ESL Council, 1992; Guyer & Peterson, 1988; Kasper, 1994a).  The academic benefit may be the result of the linguistic and cognitive skills that are developed and enhanced as the ESL student learns to read and comprehend academic content-based material in English (Ausubel, Novak, & Hanieson, 1978; Bisanz & Voss, 1981; Brinton  et al., 1989; Kasper, 1994c; Nelson & Schmid, 1989; Royer & Cable, 1975).

Academic Content and Academic Course Pairing

      Academic content within the context of academic course pairing has been shown to enhance both ESL students' reading comprehension skills and their academic performance. 

      Kasper (1994b) found that ESL students who studied a academic content-based Analytical Reading (ESL 04) course paired with an academic subject, Introductory Psychology (Psych 11) obtained significantly higher scores on final tests of English language reading skill than did ESL students who studied the same Analytical Reading (ESL 04) course in a nonthematic, nonpaired situation.  In addition to their enhanced performance on reading tests, these academically paired students performed just as well in that psychology class as their native English speaking counterparts.

      In that study, the nonpaired ESL 04 section had been a literature-based course in which the readings consisted of 2 novels, The Call of the Wild by Jack London (1903) and The Old Man and The Sea  by Ernest Hemingway (1952).  On the other hand, the text used in the paired course was The Pleasures of Psychology (Goleman & Heller, 1986), and all readings focused on various topics in the area of psychology, for example, perception, Piaget's theory of cognitive development, stress and conflict, and psychopathology.  These readings paralleled the topics students were studying in their psychology class.

      Based on the results of her study, Kasper (1994b) concluded that the enhanced reading performance of the academic paired students was the result of three factors. 

      First, paired students were better able to generalize and consolidate learning because the psychology class had provided them with multiple exposure to academic content vocabulary and linguistic structures.  Second, the parallel activities in the psychology and the ESL class offered reciprocal benefits as advance organizers and analogies for the sophisticated academic material read.  Finally, focusing on one subject area in the reading class established rich schemata which were continually activated and reinforced throughout the semester.  Paired students comprehension was thereby enhanced through the strong body of knowledge which they brought to the readings.  Kasper therefore concluded that the critical factor in the enhancement of English language reading skills was the additional exposure to complex reading material provided by the academic course pairing.

      The Kasper (1994b) study clearly demonstrated that an academically-paired content-based course improved ESL students' reading performance over a nonpaired, nonthematic, literature-based course.  However, although Kasper had found a substantial difference in reading scores between the paired and nonpaired courses she had studied, the role of academic content per se was still not clear. It was difficult to discern the critical reason for the difference in reading scores because in the paired course academic content had been confounded with academic pairing.  As a result, it was not clear how much of this difference was due to the academic content of the readings in the paired class, and how much was due to the paired students' additional exposure to the academic content in the psychology course.  Moreover, because there had been no structured academic content in the nonpaired course, it was impossible to assess the effect of such content itself on reading comprehension.

Academic Content and ESL Reading Comprehension

      The present study was conducted to clarify the role of academic content per se in enhancing ESL students' reading performance.  In this study, the nonpaired course was structured around the academic content area, psychology.  Students in the paired and the nonpaired courses both read the same selections from the text, The Pleasures of Psychology (Goleman & Heller, 1986).  Therefore, the only difference between the paired and nonpaired courses in the present study was the psychology course pairing.  Thus, the present study attempted to answer the question: Is pairing the ESL reading course with an academic content course necessary to enhance reading performance, or will enhanced reading performance result if the ESL reading course itself is structured around one academic content area? 

      In addition, the present study hoped to clarify, from both a linguistic and a cognitive perspective, any facilitative effects of academic content on reading comprehension.  The linguistic perspective focuses on the specific activities of the content-based reading course, such as advance organizers and analogies, which make the content accessible to the second language learner (Ausubel et al., 1978; Brinton et al., 1989; Kasper, 1994c; Royer & Cable, 1975).  The cognitive perspective focuses on the activation of knowledge bases, or schemata (Anderson & Pearson, 1984; Gibson & Levin, 1975; Nelson & Schmid, 1989), the development of metacognition of the reading process (Palincsar & Brown, 1984; 1986), and the processing of macrostructural (global meaning) versus microstructural (individual word) propositions (Bisanz & Voss, 1981). 

      The present study took into account both the linguistic and the cognitive perspectives on reading research.  The activities used in this study focused on developing English language skills, as well as on activating relevant schemata for the readings.  These activities included advance organizers and analogies.  The advance organizers were designed to establish appropriate background knowledge for the text to be read.  The advance organizers also made the academic content accessible to the students by introducing new vocabulary and English language structures.  Existing schemata were subsequently activated when students were asked to provide their own examples or analogies for the concepts presented in the reading. 

      In the process, students learned to construct meaning from the information stored in memory, to extract relevant information from the larger text context, and to filter out redundant or irrelevant information.  The activities used encouraged students to become active participants in their own comprehension, an important step to enhancing reading performance according to Crowder and Wagner (1992).

      In the present study, as in the Kasper (1994b) study, for each of the readings, students completed exercises that required them to summarize, question, clarify, and predict.  These four strategies have been found to increase students' metacognitive awareness of the reading process and their subsequent comprehension (Palincsar & Brown; 1986).  The exercises used asked students to use their own words to summarize a portion of the reading, to draw inferences from the reading, and to explain or clarify specific points made in the reading.  These exercises were designed to teach students appropriate strategies for extracting meaning from the reading and to transfer responsibility for comprehension to the students themselves. 

      After completing the exercise, students shared their answers with the class and received feedback from both other class members and the instructor.  Based on this feedback, students were responsible for evaluating their own answers.  By completing a series of exercises and evaluating their own answers, students were expected to develop increased metacognition of the reading process, which would subsequently lead to enhanced comprehension.  Moreover, this procedure was expected to facilitate transfer of learning by encouraging students to take responsibility for their own comprehension.  According to Weinstein (1987), one of the most powerful ways to teach transfer of learning is to provide students with practice and feedback.  Moreover, Nist and Simpson (1987) have found that when students must demonstrate their knowledge to another individual, who then provides feedback, they become involved in monitoring and evaluating, two processes critical to self-regulated learning and enhanced reading comprehension.  

      So, the present study hoped to clarify the role of academic content on ESL students' reading comprehension.  This goal would be accomplished in two ways: first, by comparing the effect of academic content versus academic pairing on comprehension, and second, by exploring the linguistic and cognitive benefits provided by academic content-based ESL reading courses.

Method

Student Population

      Kingsborough Community College, which has an ESL student population of approximately 1100 students, is located in Brooklyn, New York.  The ethnic backgrounds of the ESL students at KCC are varied and include Russian, Haitian, Hispanic, Oriental, Indian, and Arab.

      All of the ESL students are working toward, at minimum, a two-year degree, although most go on to pursue higher degrees.  Students are enrolled simultaneously in ESL and academic content courses.  Therefore, these students need instructional techniques which will facilitate their mainstream academic performance.

Academic Paired versus Academic Nonpaired ESL 04

      The study included a total of four sections of ESL Analytical Reading (ESL 04).  Two sections were academically paired with an Introductory Psychology (Psych 11) class, while two sections were nonpaired but structured around the same academic content area, psychology.

      There were 16 students in each of the four sections of ESL 04, yielding 32 students in the academic paired group and 32 students in the academic nonpaired group. Each section met for a total of 36 hours over the course of the semester.

      At the end of the semester, students in all four sections took two tests to assess their reading proficiency.  The first was the City University of New York Reading Assessment Test (CUNY RAT) and the second was the Kingsborough English Departmental Final Exam. 

      The CUNY RAT, developed by the CUNY Reading Task Force, is a multiple choice reading comprehension examination in which students have 45 minutes to answer 45 questions based on a number of short reading passages.  The KCC English Departmental Final Exam, developed by a committee of English department faculty, is a two-hour examination which requires students to write short answers to open-ended, inferential questions on a one to two page reading passage.  This test consists of one three-part multiple choice vocabulary question, seven open-ended inferential questions, and a summary question. 

      To pass ESL 04, students must pass at least one of these two tests.  The passing score on the English Departmental Final Exam is 70; the passing score on the CUNY RAT is 28.  If a student does not pass either test, he/she must repeat the course.

Readings and Activities

      Because this study was designed to compare the effect of academic course pairing versus academic content on ESL students' reading comprehension, the readings and activities used here were identical to those previously used by Kasper's (1994b) academic paired class.  In the present study, students in both paired and nonpaired groups read the same selections from the book, The Pleasures of Psychology, edited by Daniel Goleman and David Heller (1986).  As they had in the previous study, the readings focused on various topics in psychology, for example, perception, Piaget's theory of cognitive development, stress and conflict, and psychopathology.  For each of the two groups in the present study, the readings paralleled the topics that the academic paired students were studying in their psychology class (see Appendix I for a list of readings).  All pre- and postreading exercises were identical for both groups of students.

      Before each article was read, the topic treated in that article was introduced through an advance organizer or an analogy.  Overhead transparencies were used to provide graphic illustrations of the concepts presented in the articles.

      For example, before students read, "Remembering the Forgotten Art of Memory" by Thomas Scruggs and Margo Mastropieri (1992), they were shown an overhead transparency depicting the various stages in the memory system (Myers, 1990, p. 188).  The chart illustrated the sensory register, short and long term memory, and the processes of encoding, storage, and retrieval of information in memory.  These concepts were explained to the reading students by definition and through analogy.

      The Scruggs and Mastropieri article also discussed various mnemonic strategies, such as the keyword method.  To provide students with a clearer understanding of the mnemonic keyword method, the author taught them two Spanish words by showing them a picture which illustrated the imagery and acoustic associations used in the method (see Atkinson, 1975 for illustrations and a complete description of the keyword method).

      After students were introduced to its general topic in the ESL class, they were assigned to read the article at home.  After reading each article, students were given a written exercise to assess their comprehension.  Both groups completed the same exercises for each of the readings.  The instructor created a total of nine exercises, each designed to develop and enhance metacognitive skills and constructed according to the format of the KCC English Departmental Final Exam (see Appendix II for a sample of the exercises used).  Each exercise consisted of one three-part multiple-choice vocabulary question, seven open-ended inferential questions, and a summary question.

      The first exercise of the semester corresponded to the article, "Answering Questions" by Donald Norman.  To model efficient strategies for extracting appropriate information from the article, the instructor worked through this first exercise with the students.  For each question, students were asked to provide an answer and an oral explanation of how they had arrived at that answer.  To focus their attention on their own reading comprehension processes, the instructor continually asked students how and why they had chosen each answer.  They were asked to identify the part of the passage and any contextual clues they had used to answer the question. 

      Other class members were then asked if there were other possible ways to answer the question.  This was done to show the students that these open-ended questions might be answered correctly in more than one way.  Finally, the instructor provided oral feedback on the answer.  If the students had not noticed appropriate contextual clues on their own, the instructor pointed them out.  After following this modeling procedure for each question in the first exercise, the instructor gave the students a printed answer key containing her own answers to each of the questions.

      Throughout the rest of the semester, students worked alone in class to complete each of the remaining exercises.  After completing the exercise, the students were again asked to give examples of their answers to each of the questions.  They were given oral feedback on these answers both by other class members and by the instructor.  To transfer responsibility and to provide them with practice in critically reviewing and evaluating their own answers, the instructor asked the students to assign themselves points for each answer.  After the entire exercise had been reviewed as a group, students were provided with an answer key written by the instructor.  This procedure was followed for each of the remaining exercises done during the semester.

Questionnaires: Academic Nonpaired Class

      During the first week of class, students in the academic nonpaired class were asked to fill out a background questionnaire.  Students were asked whether they had ever taken ESL 04 before; whether they had taken Psychology 11 before; and how familiar they were with the subject of psychology.  These data were collected to provide insight into any effects of being an ESL 04 repeater or of having previously studied psychology.

Measures of Reading Skill

      Practice Tests

      During the course of the semester all students were given three practice tests, designed following the format of the KCC English Departmental Final Exam.  The first two practice tests were constructed by the instructor; the third practice test had been constructed by a committee of English department faculty.  Each practice test consisted of one three-part multiple-choice vocabulary question, seven open-ended inferential questions, and a summary question.  Each practice test was taken under conditions similar to those of the actual final exam; that is, students were given two hours to work, and the tests were graded by the instructor. 

      The first of these tests, the pretest, was designed to provide a comparison of the initial reading proficiency of the students in the two groups.  For useful conclusions to be drawn, it was necessary to establish that all students had started ESL 04 at the same level of proficiency.  The reading passage on the pretest was an excerpt adapted from Paul Goodman's "Growing Up Absurd" (Fox, 1983). 

      The second practice test was given at the midpoint of the semester.  This test provided an indication of any differential gains in reading proficiency between the academic paired and the academic nonpaired groups.  The reading passage on the midterm test was an excerpt adapted from "Dirt, Grime, and Cruel Crowding" by Eric Sevareid (Fox, 1983).

      The third practice test was given several days before students took the English Deparmental Final Exam.  This posttest was an actual KCC English Departmental Final Exam that was no longer in active use.

      Each of these three tests provided students with practice in answering the kinds of questions they would have on the actual final exam.  The tests also provided insights into any differences in the developing reading skills between the two groups.

      Final Reading Assessment Measures

      At the end of the semester, students in each of the four ESL 04 classes took two tests to assess their reading proficiency.  The first of these tests was the CUNY Reading Assessment Test (RAT) and the second was the KCC English Departmental Final Exam. 

Results

      Several different analyses were run to compare the reading performance of the ESL students in the academic paired group and the reading performance of the ESL students in the academic nonpaired group.  None of the analyses revealed significant differences in reading skills between the two groups.

Practice Tests

      All students in both the paired and the nonpaired group were given three practice tests, which followed the format of the KCC English Departmental Final Exam.  These practice tests were taken at the beginning (the pretest), the middle (the midterm), and the end (the posttest) of the semester.

      If we are to use any differences found in reading performance to evaluate the effect of academic content on that performance, it is necessary that we first establish that all students were reading at the same level when they entered ESL 04.  To ascertain that all students in the academic paired and the academic nonpaired groups entered ESL 04 at the same level of English language reading proficiency, a pretest was given during the first week of classes.  No significant difference was found in scores on this pretest for students in the academic paired or the academic nonpaired groups (t(62)=1.64; p  .05).  Therefore, because the students in both the academic paired and the academic nonpaired group entered ESL 04 with equivalent reading proficiency, any subsequent differences found should be due to the components of the specific ESL 04 course, rather than to any preexisting difference in English language reading skill.

      At the midpoint of the semester, students in each of the four ESL 04 classes took a midterm practice test.  Once again, there was no significant difference in scores between the academic paired and academic nonpaired students (t(62)=.039; p  .05).

      Several days before the English Departmental Final Exam, students in all four classes took a posttest.  This posttest was an actual English Departmental Final Exam that was no longer in active use.  Once again, there was no significant difference in scores between the two groups (t(62)=0.56; p  05).

      Therefore, on each of these three tests, pre-,

mid-, and post-, the reading performance of students in the academic paired groups was equivalent to that of the students in the academic nonpaired groups.  Reading performance of students in both groups was consistent throughout the semester on each of the three practice tests.

Final Reading Assessment Measures

      At the end of the semester, students took two different final assessment tests, the English Departmental Final Exam and the CUNY RAT.  Students must pass one of these tests to pass ESL 04.  The passing score on the English Departmental Final Exam was 70; the passing score on the CUNY RAT was 28.

      Table 1 presents the means and standard deviations of the scores on the English Departmental Final Exam and the CUNY RAT.

______________________________

Insert Table 1 about here

______________________________

      On the English Departmental Final Exam, no significant difference was found between the grades of the students in the academic paired group and those of the students in the academic nonpaired group (t(62)=0.65).  Differences between the two groups on the CUNY RAT were also nonsignificant (t(62)=0.12). 

      Because the students took two different final evaluation measures, but only needed to pass one to pass ESL 04, it would be interesting to know whether there was any difference between groups in the percentage of students who passed both of the exams.  Therefore, an additional analysis was run to assess this difference. No significant difference was found.

      To summarize, no significant differences were ever found between the final reading assessment scores of students in the academic paired or the academic nonpaired groups. 

Relationship of Practice Tests to Final Exam

      To determine whether students had benefitted from the three practice tests taken during the semester, correlation coefficients were computed to determine the relationship between the average grade on the three practice tests and the average grade on the English Departmental Final Exam.  Significant correlations were found for the scores in both groups.  The correlation coefficient for the academic paired group was r(32)=+0.57; p  .02.  The correlation coefficient for the academic nonpaired group was r(32)=+0.65; p  .01.

      The significant relationship between the average scores students obtained on the three practice tests and the scores they obtained on the final exam means that practice test scores may be used to predict performance on the final exam.  The practice tests appear to have provided students with enhanced metacognition of the reading task, thereby facilitating performance on the actual final exam.

Effect of Having Studied Psychology 11: Academic Nonpaired

      An analysis was run to determine whether the academic nonpaired students who had studied Psych 11 would have an advantage over the academic nonpaired students who had never studied it.  This analysis compared the CUNY RAT and English Departmental Final Exam scores of both groups of students.  Both comparisons were nonsignificant, so having studied Psych 11 previously did not affect performance on either of the two final reading assessment measures.

Student Feedback

      At the end of the semester, ESL 04 students in  the academic paired group and the academic nonpaired group were asked to complete questionnaires which asked for their feedback on the reading course.

      Academic Paired

      Examples of the questions academic paired students were asked include: (1) do you think pairing ESL 04 and Psych 11 is a good idea?; (2) did the English course help you to work on your psychology?; (3) did you find the amount of material in both ESL and Psych classes overwhelming; and (4) would you recommend this type of program with paired courses to a friend?  Students were asked to answer "yes" or "no" to the questions and to provide a reason for each of their answers.  100% of the students answered "yes" to questions 1, 2, and 4.  67% of the students answered "no" to question 3.  So, all of the students thought that the experience was worthwhile, and two-thirds of them felt that the workload was manageable.

      In addition, students were asked whether they felt the articles had (a) helped them to understand the psychology topics, (b) helped to prepared them to take the English Departmental Final Exam, and (c) helped to prepare them to take the CUNY RAT.  100% answered "yes" to questions (a) and (b), and 80% answered "yes" to question (c).  Therefore, students believed that the academic paired course provided both academic and linguistic benefits.

      Academic Nonpaired

      Because a academic nonpaired approach was a new method of teaching ESL 04, student feedback would be an important consideration in the design of future ESL 04 academic nonpaired courses.  Students were asked to answer "yes" or "no" to each question and to provide a reason for each of their answers.  Examples of some questions are: (1)did you enjoy reading about one subject area all semester?; (2)do you think the readings and the exercises done were helpful to preparing you for the English Departmental Final Exam and the CUNY RAT?.  94% of the students answered "yes" to question 1; 100% answered "yes" to question 2.

      Students were also asked, "how confident are you that you will pass the Departmental Final/the CUNY RAT?"  97% expressed confidence that they would pass the Departmental Final; while 93% expressed confidence that they would pass the CUNY RAT. 

      In addition, students who had not taken Psychology 11 before were asked, "has this ESL 04 academic contentbased reading course made you interested in taking Psychology 11?"  100% of these students answered "yes" to this question.  There was one student who had previously dropped Psychology 11 because he felt it was too difficult for him.  He said that after taking this ESL 04 reading course, he would register for Psych 11 again because he now realized it was not as difficult as he had thought. 

      Other student responses indicated that this reading course had sparked interest in studying psychology because the course gave an idea of what psychology is all about and helped develop more understanding of psychological readings.  So, like their academic paired counterparts, students in the academic nonpaired class enjoyed reading about psychology and felt that the course had benefitted them both linguistically and, to some extent, academically.

Discussion

      What conclusions about the effect of academic content on ESL students' reading performance may be drawn from the results of the present study?  What light has this study shed on the linguistic and cognitive explanations for this effect? 

Academic Content versus Academic Course Pairing

      Both reading assessment measures yielded equivalent performance for the academic nonpaired students and the academic paired students.  Based on the conclusions of Kasper's (1994b) study, one might expect the academic paired students to have achieved higher scores because they were provided with multiple exposure to the content material in the psychology class.  In that study, Kasper theorized that academic course pairing enhances reading performance in three ways.  The first is through multiple instructional contexts, leading to greater generalization and consolidation of learning.  The second is through the reciprocal benefits of activities as advance organizers and analogies in the psychology and the ESL class, respectively.  The third is through the establishment of elaborate schemata which were continually activated throughout the semester.

      In the present study, because students in the academic nonpaired group performed just as well on tests of reading comprehension as did students in the academic paired group, it seems that multiple instructional contexts are not necessary to enhance reading performance.  What does appear to be necessary to produce enhanced reading comprehension is an understanding of how to approach the reading of different types of passages and how to construct meaning by extracting appropriate information from the passage. 

      The results of the present study suggest that the academic content of the readings, rather than the multiple exposure to that content provided by course pairings, is the critical factor in enhancing ESL students' reading comprehension.  Readings of academic content facilitate performance for three major reasons.  Increased reading skill results because academic content helps students construct schemata, develop efficient comprehension strategies, and increase metacognition of the reading process.

Construction of Schemata

      By introducing new information into the knowledge system, academic content helps the reader construct schemata.  That is, each time students read a selection containing academic content, they learn something new.  Assimilating this new information requires the formation of new schemata and/or the accomodation of existing schemata.  According to Gibson and Levin (1975), as  schemata are developed and elaborated, the scope of the cognitive structure is increased.  As a result, it becomes easier to understand what is read because the knowledge necessary for comprehension is more likely to be present in the cognitive system.  The higher the level of students' topical knowledge, the better able they are to process macropropositions and to construct an integrated representation of the text in terms of its overall meaning (Bisanz & Voss, 1981).

Comprehension Strategies

      In addition, focusing on academic content teaches students efficient comprehension strategies, thereby facilitating the overall reading process.  A critical factor in enhancing ESL students' reading comprehension appears to be teaching them to focus on macrostructural propositions (meaning) rather than on microstructural propositions (individual words).  Comprehension exercises that develop this ability help to foster enhanced reading comprehension.

      The exercises used in the present study were designed to teach the students to use more effective reading comprehension strategies.  The open-ended questions in the exercises may have tapped deeper levels of information processing (Carrell, Pharis, & Liberto, 1989).  Moreover, the sequence of instruction facilitated students' independent use of efficient reading strategies by preceding that independent use by direct explanation and guided practice.  Shih (1992) suggests that "the most effective way to promote student use of a strategy is by following a sequence of instruction in which independent use is preceded by direct explanation and guided practice" (p. 300).   By following such an instructional sequence, students in the present study learned how to extract information from the passages and how and when to use various strategies.  By evaluating their own answers, they also learned to determine how well they had used each strategy. 

      Students in this study continuously practiced extracting information from different passages dealing with various topics in psychology.  The passages contained difficult vocabulary words, many of which were unfamiliar. 

      Through the instructional techniques used, students learned not to focus on individual words, but to try to grasp the overall meaning of each section of the article.  Both the pre- and post-reading exercises directed their attention to the information they would need to understand the article.  The advance organizers and analogies activated relevant schemata for the topic.  The instructional procedure, including the exercises and practice tests, gradually enabled students to assume greater responsibility for their comprehension.  As students moved throughout the semester, they became increasingly able to articulate the steps they had followed as they read the articles and answered the questions in the exercises.

      Academic Content versus Literature

      Is following an instructional sequence of feedback and practice with reading comprehension exercises sufficient to lead to enhanced reading performance?  The results of the present study and those of Kasper's (1994b) study suggest that it is not.  Although Kasper's students in the nonthematic, nonpaired literature-based class had also completed reading comprehension exercises similar in design to those completed by her academically paired students and by the students in the two academic content-based groups in this study, her nonthematic literature students had obtained significantly lower scores on the English Departmental Final Exam.  The questions on the literature-based comprehension exercises were designed following the format of the English Departmental Final Exam, and the same instructional technique of direct explanation, guided practice, and independent application was used. 

      Therefore if teaching students effective reading comprehension strategies were enough to produce enhanced reading skill, Kasper's literature-based reading students should have performed as well on measures of reading proficiency as the academic content-based reading students both in her study and in the present study.   Clearly however, the type of material ESL students read plays a role in the development of their English language reading skills.

      Why then does academic content enhance reading performance over literature?  In terms of schema theory, it may be that academic content develops more elaborate schemata than literature does.  Literature tends to be interpreted within the framework of our existing experience and knowledge.  Thus, the comprehension of literature may require fewer interacting knowledge sources.  Put more simply, we can understand the gist of literature without adding much new information to our existing knowledge base.  The comprehension of literature is more subjective, so less accomodation of schemata may be necessary.

      It is also possible that the nature of the reading passage on the English Departmental Final Exam makes academic content a more appropriate way to prepare students for this test.  The reading passage on the Final Exam consists of an excerpt from a topical article, on for example, electronic monitoring or fast foods.  Readings of academic content are closer in nature to these types of passages than are works of literature.  So, in terms of the readings, students in a academic, content-based course are provided more practice with the type of material they will find on the Final Exam.

Increased Metacognition

      A content-based, strategy-oriented approach to reading instruction increases students' metacognitive control and allows them to develop a new schema for how to understand what they read.  This increased metacognition allows students to take charge of their own learning and to become spontaneous strategy users.  Their schema for the reading process contains the knowledge that the strategies of summarizing, questioning, clarifying, and predicting (Palincsar & Brown, 1986) work to enable the reader to construct meaning from the text, whatever the text may be.  Whenever and whatever they read, they call up this schema for reading and then apply the strategies to the text.  This leads to overall enhanced reading comprehension and to transfer of skill as found by Nelson and Schmid (1989).

      When asked whether they thought academic reading courses were helpful to their English language reading development, the ESL students in this study offered their own insights into the metacognitive benefits gained from the content-based, strategy-oriented reading approach.  They said that the psychology readings were interesting and complex, and that the comprehension exercises had taught them how to read, absorb, analyze, and organize information.  They believed that they had learned how to read and understand complex academic material, and that this skill would be useful in their mainstream academic courses.

Conclusion

      The present study makes a strong case for teaching academic content-based ESL reading courses.  These courses not only help ESL students to become better readers of English and encourage them to pursue further study of the content area in mainstream academic classes, but the courses also enable ESL students to perform at the level of native speakers in those academic classes.  Academic content may be introduced into any ESL reading course, be it paired or not.  Because of their linguistic and cognitive benefits and their demonstrated success in enhancing reading skills, academic content-based reading courses represent a worthwhile addition to the ESL curriculum.