![]()
PART II
CognitiveāElaborative Approaches to Peer Learning: Overview
The six chapters in this part describe a variety of techniques that depend to a greater or lesser extent on cognitive elaboration theory (Slavin, 1995). Such approaches are based on general information processing theory and suggest that in order to remember information and connect it to existing knowledge, a student must restructure the information or elaborate on the information (Wittrock, 1978, 1990). The chapters also draw on the cognitive-developmental theories of Piaget and Vygotsky described in Part I of this book. In addition, many of the chapters in this part are strongly influenced by sociocultural perspectives on learning and often reflect cognitive-constructivist perspectives.
The chapters in this Part II describe specific techniques or models of collaboration. Three of the chapters (chaps. 3, 4, & 5) specifically address the quality of discourse within a peer learning group and issues related to providing effective support for productive interaction. Two additional chapters (chaps. 6 & 7) extend this discussion. The last chapter in this part provides a commentary on these chapters. Because the various chapters included here describe specific techniques, we thought it important to end this part of the book with a chapter that provided some integration of the other chapters. The final chapter by Derry draws attention to the underlying themes or ideas that are common to these chapters.
![]()
CHAPTER THREE
Evolution of Discourse During Cross-Age Tutoring
NATALIE K. PERSON
Rhodes College
ARTHUR G. GRAESSER
It has been well documented that one-to-one tutoring is an effective method of instruction. Reported effect sizes have ranged from 0.4 to 2.3 standard deviation units when tutored students are compared to classroom instruction or other control groups (Anderson, Corbett, Koedinger, & Pelletier, 1995; Bloom, 1984; Cohen, Kulik, & Kulik, 1982; Mohan, 1972). These effect sizes are quite surprising considering that a normal tutor typically has minimal tutoring experience, lacks expert domain knowledge, and has received no training in sophisticated tutoring techniques (Fitz-Gibbon, 1977; Graesser & Person, 1994). Although the majority of studies that report the benefits of tutoring primarily focused on outcomes, some of the more contemporary research is aimed at understanding the tutoring process (Fox, 1991, 1993; Graesser, Bowers, Hacker, & Person, in press; Graesser & Person, 1994; Graesser, Person, & Magliano, 1995; Lepper, Aspinwall, Mumme, & Chabay, 1990; Merrill, Reiser, Ranney, & Trafton, 1992; Person, Graesser, Magliano, & Kreuz, 1994; Putnam, 1987). Many of these researchers have attempted to understand the effectiveness of normal, unskilled tutors by systematically analyzing the collaborative dialogue that occurs between tutors and students.
The purpose of this chapter is to illustrate the pedagogical mechanisms that facilitate learning during normal one-to-one tutoring interactions. We believe that the advantages that one-to-one tutoring has over other learning methods are best understood by analyzing the contributions that tutors and students make while attempting to answer a question or solve a problem. In a number of previous publications and manuscripts, we documented the pedagogical tactics and strategies that tutors frequently implement during these interactions. We also provided substantial evidence that tutors rarely employ sophisticated tutoring techniques that are often incorporated into structured tutoring facilities in school systems and intelligent tutoring systems on computers (Graesser et al., in press; Graesser & Person, 1994; Graesser et al., 1995; Person, 1994; Person et al., 1994). In this chapter, we illustrate some of our previous claims about tutor strategies via actual tutoring dialogue excerpts. The chapter contains four sections. We first briefly describe the two tutoring samples that we used in our analyses. We then discuss the five-step dialogue frame that is prevalent in tutoring sessions but not in classrooms. A general understanding of the five-step frame is important because many of our subsequent arguments are housed within this framework. The third section provides a brief overview of sophisticated strategies that are rarely implemented by normal tutors. The last section illustrates the pedagogical strategies and dialogue moves that normal tutors did use in our tutoring samples.
TWO TUTORING SAMPLES
Tutoring protocols were collected from two naturalistic tutoring samples. The students in the first sample consisted of 27 undergraduates enrolled in a psychology research methods course who were required to participate in the tutoring sessions in order to fulfill a course requirement. The instructor of the research methods course identified six topics that are typically difficult for students to master. These topics included (a) operationally defining variables, (b) interpreting graphs, (c) statistics, (d) designing an experiment to test a hypothesis, (e) factorial designs, and (f ) interpreting interactions. A counterbalancing scheme ensured that each student attended two different tutoring sessions with two different tutors.
The tutors for the research methods sample were three psychology graduate students who were recommended by a faculty member and were considered to be competent in the area of research methodology. Each of the tutors had some prior tutoring experience but not in the area of research methods.
The students in the second sample consisted of 13 seventh graders who were having difficulty in algebra and who were recommended for tutoring by their algebra instructors. They were tutored on topics such as fractions, exponents, positive and negative numbers, and algebra word problems. This sample included all of the algebra tutoring that took place at this particular middle school during a 1-month period. Therefore, it should be noted that this sample is representative of the tutoring that takes place in school settings.
The tutors for the algebra sample were 10 high school students who had performed well in previous mathematics courses. None of the tutors in these two samples had extensive tutoring experience (approximately 9 hours each), had ever received formal training in tutoring strategies, or could be considered experts in their respective domains. Therefore, these tutors were typical of the tutors who typically work with students in school settings. The algebra tutors all attended the same school as the seventh-grade algebra students and were roughly 3 years older than the tutees.
The tutors and students in the two samples, however, should not be considered peers in the sense that their level of expertise is fundamentally equal. Other researchers have investigated learning situations in which two peers with roughly equal knowledge collaborate to solve a complex task (OāDonnell, Dansereau, Hythecker, et al., 1988; OāDonnell et al., 1990; Rogoff, 1990). Our tutors are not peers in this sense.
All tutoring sessions were videotaped and transcribed. The research methods sample had 44 hours of tutoring, whereas the algebra sample had 22. Some of the sessions in the research methods sample could not be transcribed because of poor sound. A full account of our data collection, transcribing procedures, coding methods, and statistical analyses can be found in several previous articles (Graesser & Person, 1994; Graesser et al., 1995; Person et al., 1994; Person, Kreuz, Zwaan, & Graesser, 1995). Our previous analyses indicated virtually no differences between the two samples in terms of the dialogue patterns, pedagogical strategies, and tactics that were employed by the tutors. Therefore, we do not attempt to differentiate these two samples in our illustration of such patterns, tactics, and strategies.
THE FIVE-STEP DIALOGUE FRAME
Some researchers have characterized classroom interactions in terms of a three-step dialogue frame. The sequence of these three steps involves: (1) the teacherās request for information from a student, (2) the studentās answer, and (3) the teacherās evaluation of the studentās contribution. Mehan (1979) referred to these three steps as initiation, response, and evaluation, whereas Sinclair and Coulthart (1975) referred to them as question, answer, and evaluation. Regardless of the labels (IRE or QAE), neither of these three-step frames requires extensive input from the student.
The dialogue frame that is prevalent in tutoring consists of five steps rather than three. Our previous publications provide detailed descriptions of the five steps (Graesser & Person, 1994; Graesser et al., in press; Graesser et al., 1995). A brief description of the five steps follows:
1. Tutor asks a question (or alternatively provides a problem for the student to solve).
2. Student answers the question.
3. Tutor gives feedback on the answer.
4. Tutor and student collaboratively improve the quality of answer.
5. Tutor assesses studentās understanding of the answer.
The five-step dialogue frame is illustrated in the following exchange from a tutoring session on the topic of factorial designs (research method).
Example 1
Step 1 | 1:1 | TUTOR: | So, how many F scores would be computed? |
Step 2 | 1:2 | STUDENT: | Three. |
Step 3 | 1:3 | TUTOR: | Three [agreeing with the student]. |
Step 4 | 1:4 | TUTOR: | And what numbers [referring to a matrix of cell means] would you use? |
| 1:5 | STUDENT: | You would do one for humor [one of the independent variables]. |
| 1:6 | TUTOR: | And what does that tell you? |
| 1:7 | STUDENT: | Iām not sure [laughs]. |
| 1:8 | TUTOR: | OK, why do you do an F score? What is an F score? |
| 1:9 | STUDENT: | To see the size, uh, significance? |
| 1:10 | TUTOR: | The size of significance. |
| 1:11 | STUDENT: | The size of significance. |
| 1:12 | TUTOR: | Right, how statistically significant a variable is. |
| 1:13 | STUDENT: | Right. |
| 1:14 | TUTOR: | So, you are right, you would have three [F scores]: one for caffeine, one for humor, and one for . . . ? |
| 1:15 | STUDENT: | The scores . . . from caffeine and humor. |
| 1:16 | TUTOR: | Interaction, the interaction of the two, right? |
| 1:17 | STUDENT: | Um hmm. |
| 1:18 | TUTOR: | [Explains independence of main effects]. |
Step 5 | 1:19 | TUTOR: | Do you see what Iām saying? |
| 1:20 | STUDENT: | Um hmm. |
The nature of this dialogue is quite different from what typically occurs in classroom interactions. Notice that Step 4 consists of multiple turns that allows the tutor and student to elaborate on ideas raised in the earlier steps. It would be quite difficult for a teacher in a classroom to craft such an elaborative dialogue that would meet the needs...