For more than 100 years, educational, developmental, and social psychologists have researched and developed theoretical explanations for how students learn, remember, and, as a result, know. Several of the most important theories have originated from attempts to understand and explain child development.
Early cognitive theorists (e.g., Piaget, 1926) were interested in the ways cognition developed – the faculty of knowing. Piaget believed that young children’s thoughts influenced their language and viewed children’s thoughts as important to the outcome of the developmental process. Piaget focused his interest on how a child’s interaction with the environment leads to the progressive development of cognitive abilities. He further acknowledged that interaction with other children or peers stimulates children to become more aware of the perspectives of others. Unfortunately, Piaget was not able to explain children’s individual differences, factors that account for these differences, or ways to promote intellectual development.
Behaviorist theorists focused on learning as a function of items such as reinforcement and punishment but also strongly promoted the importance of the environment as the principal force shaping development. Early behaviorists (Pavlov, 1927; Skinner, 1953, 1957, 1961, 1968; Watson, 1925) focused on the child’s immediate behavior and the environmental forces (conditioning – both classical and operant) that affect the behavior. Behaviorists contend development can be best understood through the analysis of specific behaviors, the circumstances leading to them, and their consequences. In addition, behaviorist theories are poorly suited to explain higher mental processes (e.g., thinking, feeling, analyzing, problem solving, evaluating, etc).
In order to overcome apparent deficiencies inherent in previous explanations of how humans learn, remember, and consequently know, many researchers have attempted to combine the best elements from various theories. A growing number of contemporary educational researchers as well as developmental psychologists are profoundly influenced by social cognitive theories. These theories of social cognitivism blend the best of the behaviorist tenets (recognizing that learning involves models of various kinds that act as social influences on the child) with the best of the cognitive tenets (recognizing the importance of a child’s ability to reason, to uncover cause-and-effect relationships, and to anticipate the outcomes of behavior). Bandura’s (1977, 1981, 1986) observational theory originally stemmed directly from a behavioristic orientation but has become progressively more cognitive and concerned with knowing, understanding, thinking and mental processes.
Bandura’s theory is an application of operant conditioning based on the effects of imitation. He describes four processes involved in observational learning: (a) attentional processes in which the observer attends to important aspects of the model’s behaviors; (b) retentional processes in which the observer mentally represents, in images or words, and stores in memory what has been observed; (c) production processes that make possible the performance of the observed behavior and; (d) motivational processes that lead to actual performance rather than acquiring behavior without performance. The three manifestations of observational learning are (1) the acquisition of novel responses (the modeling effect); (2) the inhibition or disinhibition of deviant responses (the inhibitory or disinhibitory effect), and (3) the encouragement of behaviors that are neither novel nor deviant but that are directly related to those of a model (the eliciting effect).
The question of how humans learn, remember, and know was being researched and investigated in 1902 and it is being researched and investigated in 2002. Research journals publish articles that investigate and produce evidence regarding new and potentially innovative instructional strategies (individualistic vs. collaborative; competitive vs. cooperative), the influence of new instructional technologies (computer-based instruction vs. face-to-face instruction) on the facilitation of knowledge, and the ability of students to learn without adequate training (a.k.a., in spite of their teachers).
The theory and research that educational, developmental, and social psychologists have developed over more than a century should not be abandoned. Instead, one should assimilate, extend, and test current applications of competing theoretical explanations for how humans learn, remember, and, consequently, know. For example, if one takes Vygotsky’s (1962, 1978) proposal that social interaction through language and with adults and cognitively advanced children facilitates cognitive development and combines it with Bandura’s (1971, 1981, 1986) observational (social) learning theory, then one is better able to compare individuals receiving computer-based instruction with trained cooperative learners who also receive computer-based instruction. This comparison allows a prediction about the potential benefits of trained social interaction to facilitate cognitive learning.
Social Interdependence Theory
Social interdependence exists when individuals share similar purposes for a task; success is also shared and relies on others' actions (Johnson & Johnson, 1996). Deutsch’s (1949b) theory of cooperation and competition provided a focal point for the formulation of social interdependence theories in the mid- to late 1900s. Johnson and Johnson stated that Deutsch’s theory of cooperation and competition, established on the work of Lewin, was the focal point of more recent studies of cooperative learning (1991).
Johnson and Johnson (1996) credited social interdependence theory’s origin to Kurt Lewin and the Gestalt school of psychology. Gestalt psychology studies how people view and comprehend the relation of the whole to the parts that make up that whole (Winn & Snyder, 1996). The ideas of the Gestalt school of psychology concerning the structure of social groups have influenced the use of cooperative learning. It is no longer necessary, as Lewin (1947) claimed it once was, for researchers to argue the very existence of a group within a societal structure.
Deutsch (1949a; 1949b) extended Lewin’s theories and formed a theory of cooperation and competition. Deutsch recognized three types of social interdependence: positive, negative, and an absence of social interdependence. Under Deutsch's view, the type of interdependence within a situation determined an individual’s interactions, which subsequently determined outcomes (Johnson & Johnson, 1996).
Positive interdependence is seen as promoting interaction among group members. Negative interdependence is seen as detracting from group interaction. The condition of no interdependence is the result of an absence of group interaction. The idea of interdependence among group members has influenced most of those investigating questions centered on cooperative learning, most notably, D. Johnson and R. Johnson as well as Slavin.
Cooperative Learning Theory
Three of the most prominent researchers of cooperative learning are Slavin and Johnson and Johnson. Analyzing their work lends a unique perspective to formal, structured cooperative learning. Slavin (1994) believed all cooperative learning methods had certain shared central ideas. In cooperative learning, students work together to learn and are responsible for one another’s learning as well as individual learning. Johnson and Johnson (1996) defined cooperative learning as the pedagogical use of small groups of two or more students who work together to maximize their own and each other’s learning.
Johnson and Johnson identify four types of cooperative learning. In formal cooperative learning, students work together to achieve shared learning goals and jointly complete educational tasks (Johnson & Johnson, 1996). Formal cooperative learning also proposes that teachers should tell the students the objectives for the lesson, make several preinstructional planning decisions, clearly explain the task to the students (as well as the need for positive interdependence), monitor student learning, intervene to provide assistance, and, finally, evaluate students’ learning and help the students evaluate their own learning.
In the second type, informal cooperative learning, teachers assign students to work together to achieve a joint learning goal. The learning groups are temporary and meet for small periods of time (Johnson & Johnson, 1996).
The third type of cooperative learning is the cooperative base group which consists of stable membership over a long time period and group members of mixed ability (Johnson & Johnson, 1996). Base groups are established to support peer group members throughout the academic year with the goals of making academic progress along with positive cognitive and social development. Cooperative base groups typically are seen in formal cooperative learning settings.
The fourth type of cooperative learning is academic controversy, in which one student's thoughts, ideas, or other knowledge formulations are incompatible with those of a second student and the two seek to reach an agreement (Johnson & Johnson, 1996). Academic controversy also can consist of formal or informal cooperative learning.
Other significant cooperative learning methods articulated by Slavin (1994) include Student Teams-Achievement Divisions (STAD), Teams-Games-Tournament (TGT), Team Assisted Individualization (TAI), Jigsaw, and group investigation. STAD consists of five major components. These parts are class presentations, team activities, quizzes, individual improvement evaluations, and team recognition. TGT is similar to the STAD approach to cooperative learning with one significant instructional difference: TGT uses academic tournaments instead of quizzes and individual improvement scores. TAI is focused on the individualization of mathematics instruction and uses a specialized set of curriculum materials. Jigsaw is a form of instruction using cooperative learning techniques and is most appropriate when curriculum material is in written narrative form. Students working in Jigsaw, STAD, TAI, and TGT all work in heterogeneous teams.
Group investigation is a form of cooperative learning that dates back to John Dewey (1974). Student success using group investigation requires prior training in communication and social skills. For decades, many researchers have investigated the impact of students working together in instructional situations.
Reviews of Cooperative Learning Without Computers
The role of cooperative, competitive, and individual effort on achievement and performance has been investigated in the laboratory since the 1920s (Maller, 1929) . Deutsch (1949b) presented an influential theory of cooperation and competition concerning their effects on small group functioning. The theory proposed an effect of strength of group membership and degree of unity for the functioning of small groups. In a subsequent article, Deutsch (1949a) reported the results of an experimental study generally supporting his theory of cooperation and competition. The study noted increased coordination of efforts between group members, greater diversity in amount of contributions per member, higher levels of subdivision of activity, and multiple levels of communication improvement between group members.
Slavin (1980) reviewed 28 cooperative learning field projects that compared the effectiveness of cooperative and traditional learning strategies. A primary conclusion drawn by Slavin was that cooperative learning techniques were no worse than traditional techniques for academic achievement. In most cases, they were significantly better. Johnson, Maryuama, Johnson, Nelson and Skon (1981) reviewed 122 studies comparing the relative effectiveness of cooperative groupings, competitive groupings, and individual goal structures. They drew the following conclusions from the statistical meta-analysis: (a) cooperation was considerably more effective than interpersonal competition and individualistic efforts, (b) cooperation with intergroup competition was also superior to interpersonal competition and individualistic efforts, and (c) no significant difference existed between interpersonal competitive and individualistic efforts. Slavin (1983) conducted another review of research on the achievement effects of cooperative learning strategies. Performance was higher in cooperative learning groups versus control groups for 29 of the 46 studies included in his review. Slavin concluded that enhancement of student achievement could occur when cooperative methods that used group study and group rewards for individual learning were incorporated into instruction.
The use of personal computers in classrooms to assist students in achieving cooperative learning goals has increased since the 1980s (Slavin, 1980). Initially, a concern existed that the introduction of computers would adversely affect the development of students’ social and communication skills. This concern has not been realized as students often work together with computers. Increased communication and social activity has actually occurred in many instructional settings that use computers. In some cases the instructor planned for group work. In many instances using a group work strategy increased student access to the few computers available.
Reviews of Cooperative Learning With Computers
Webb (1987) reviewed the literature concerning peer interaction and learning with computers in groups. The goals of this review were to determine the following: (a) the pros and cons of group work for learning, (b) the types of verbal interactions that occur when small groups of students work at a computer, and (c) the types of interactions that were beneficial or detrimental to learning. She concluded that group work with computers was a feasible and capable way to learn. According to Webb, it was possible to design group-learning settings that benefited most students (Webb, 1987) . Hooper and Hannafin (1991), Rysavy and Sales (1991), and Simpson (1986) also published literature reviews or discussions of cooperative learning and computer-based instruction. All three articles found numerous positive influences of computers and cooperative learning.
Summary Guide to the Literature: Cooperative Learning and Computer-Based Instruction
A comprehensive study of the literature revealed significant effects for more than 134 experimental comparisons. Appendix A contains a summary of these comparisons.
Computers and Cooperative Learning: Individual and Group Variables
In their writings on the processes of computer-assisted cooperative learning, D. Johnson and R. Johnson (1996) identify several key functions of group learning. These are positive interdependence, equality of participation, co-construction of ideas, giving and receiving help, promotive interaction, division of labor and conflict and controversy. Positive interdependence exists within a group when all members of a group believe that success is a joint endeavor. Group members are required to work together to achieve mutual, not individual, success. Equality of participation means that members of the group and the group as a whole are accountable. Group accountability exists when group performance is measured according to a particular criterion as opposed to individual accountability which exists when the performance of individuals is assessed, and results of that assessment are returned to the group for comparison against a measure of performance. Group members are held responsible for contributing an equitable share to the success of the group and the group typically reviews individual assessments. Promotive interaction occurs when members of a group encourage and facilitate other members’ tasks in order to reach the goals of the group. Giving and receiving help occurs as group members aid each other in the pursuit of their mutual task. Division of labor is the process of dividing the group task into manageable sub-tasks for group members. Conflict and controversy, also a part of the cooperative process, occur as group members negotiate over the group task. Co-construction of ideas occurs as the learning occurs: Group members pursue a task and create a jointly constructed idea of their response to the instructional task.
Roschelle and Teasley (1995) examined the construction of shared knowledge in joint computerized problem-solving space. The computerized problem-solving space created for their research was viewed as essential to the completion of the cooperative task. The cooperative task was to view a graphical simulation of the concepts of velocity and acceleration. Researchers observed that the use of the computer created a space for clarifying discussions between group members. Cooperative computerized activities were seen as a means of resolving conflict or impasses. The specially designed interface was a device that both invited and constrained students’ interpretations of dyad members' communications.
Rubtsov (1992) examined joint action at the computer. He constructed several principles following his research with a coordinated computerized task between dyad members. Several conclusions were derived from this research regarding the stages of joint action organization. During the initial stage of the research, participants were concerned with the immediate external consequences of their action. Gradually, students realized the way that their individual actions were distributed in relation to each another and the level of coordination between the two. Different levels of achievement were related to different levels of participant awareness regarding the relation between the structure of their coordinated action and the structure of the corresponding outcome.
Scardamalia, Bereiter, Mclean, Swallow and Woodruff (1989) examined several instances of joint construction of ideas in their investigation of Computer Supported Intentional Learning Environments (CSILE). Students who were using a CSILE were provided a means to build a collective networked database examining a particular educational idea. Scardamalia et al. (1989) maintained that the use of a networked CSILE aided in student construction of ideas through the use of direct contributions by students as they prepared for class contributions. Students were aided in the acquisition of higher-order executive control of the learning process. The CSILE endeavor was one of the most extensively researched projects (Oshima, 1989; Scardamalia, 1989; Scardamalia et al., 1989; Scardamalia & Bereitner, 1991; Scardamalia et al., 1992; Scardamalia et al., 1994; Oshima, Bereitner, & Scardamalia, 1995; Oshima, Scardamalia, & Bereitner, 1996; Scardamalia & Bereitner, 1996) in cooperative computer-based instruction in the recent decade.
Researchers have found the area of conflict and controversy among members of a cooperative group worthy of considerable attention. Howe, Tolmie, Anderson and Mackenzie (1992) examined the role of group interaction in computer-supported teaching. In their summary of multiple studies, Howe, Tolmie and Mackenzie (1995) maintained a general principle that software that supports computerized group work should require students to explicitly state and agree on their joint predictions. Agreement, as the Piagetian perspective provides, is only arrived at in a group situation by means of explored conflict. Joiner (1995) examined the dialog of students engaged in a cooperative group task. His research showed that modeled predictions of student interactions resolving conflict did exist. This research provided the basis for modeling conflict-generating situations.
Clements and Nastasi (1988) have examined the role of conflict and controversy in their program of research into Logo-based educational environments. Logo is a computer-based programming language originated by Papert of the Massachusetts Institute of Technology (1993). In one representative study, Clements and Nastasi (1988) found significant group differences among groups of students using Logo for conflict resolution, rule determination, and self-directed work. They did not find large differences between students using traditional computer-aided instruction and students using Logo-based instruction. These researchers did, however, find that both Logo- and computer-aided instructional environments encouraged interaction and decision-making skills among subjects.
Studies organized by D. Johnson and R. Johnson were extremely influential in the investigation of cooperative learning strategies. In an article summarizing several pieces of research (1993), they stated that cooperative learning promoted greater levels of oral discussion of curriculum material, higher achievement, and frequent use of higher-order reasoning strategies when compared with competitive and individualistic learning.
Numerous researchers have examined different aspects of group processes, both in the traditional study of cooperative learning situations and in the newer study of computer-based cooperative learning. Webb and Lewis (1988) examined several aspects of help-giving behavior during their program of research and discovered that several factors of student discussions reflected positive correlations. These correlated behaviors included giving explanations and input suggestions, receiving responses to questions, and receiving input suggestions in a group programming exercise.
Researchers did not find positive correlations in the verbalization of help-giving behavior. Jackson, Fletcher, and Messer (1992) found no significant effect of verbalization on performance. Their experimental findings, however, were tempered by a highly visual component of the experimental task. Laurillard (1992) found that non-canonical display diagrams, direct manipulation interfaces, and time-based constraints all serve to enhance reflective dialogue in a cooperative computerized instructional situation. These results reflect a realization few researchers have examined: The design of the interface for the cooperative computerized task was often as important as the design of the task itself.
Hooper (1992) extensively researched behaviors surrounding the cooperative behaviors of giving and receiving help while using a computer between higher- and lower-ability students. He found that high-ability students generated and received significantly more help in groups of similar ability levels than when placed in groups of mixed ability levels. Hooper's findings also suggested that, when grouped heterogeneously, high-ability students received lower amounts of stimulation in conversation with lower-ability students. Sherman (1994) examined help-giving behavior. He found that dyads using a cued HyperCard version of a treatment showed significantly more helping behaviors than those who used a noncued version of the same treatment. Several researchers noted that the use of prompting cues in programs has potentially beneficial effects in increasing helping behaviors.
Although several researchers examined behaviors surrounding the giving and receiving of help in cooperative computer use, a need for research still exists in this area. In a summary of past research and call for future studies, Light and Blaye (1990) emphasized the need for researchers to be able to specify the ways in which children solicit and gain help from those around them while working at the computer. This knowledge of socially mediated help should then be used to design help systems for computers. The research literature on computerized cooperative learning has focused primarily on helping behaviors. Equality of participation, social loafing, and division of labor are all areas that have been neglected in the literature.
More than 20 researchers have examined gender as an experimental factor. Underwood, McCaffrey and Underwood (1990) found that single gender pairs of elementary school children showed improvement in task performance when compared to the same children working individually. Mixed gender pairs showed no relative improvement. In a subsequent study, researchers found that girls tended to cooperate even when instructed not to organize or create roles for each other. Mixed pairs tended not to cooperate, even when instructed to share task work. Boys did not cooperate, unless so instructed specifically, after which their performance improved (Underwood, Jindal, & Underwood, 1994).
Tolmie and Howe (1993) also examined the question of gender differences between 82 twelve-to-fifteen-year-old members of cooperative computer-using pairs and found a convergence to the norms of the other gender in mixed groups. Males were observed to display more behaviors similar to those of females and females displayed more masculine behaviors, if the group members were aware of norms for each gender. Herschel (1994) examined group gender composition using a networked group support system environment. Herschel studied 61 groups consisting of 269 university students. In his study, Herschel found no significant differences between gender-based idea generation using the networked group environment. This study provided evidence of the leveling effect often seen within a networked environment. When groups interacted on an electronic basis using a networked environment, gender differences that typically occurred in a face-to-face setting were minimized. This finding contrasted with the Underwood studies cited earlier. The major difference between the findings of Herschel (1994) and Underwood et al. (1994) was the presence or absence of nonverbal communication factors. Age-related factors also accounted for differences between the two studies: Underwood (1994) examined children as research subjects and Herschel (1994) examined university students.
Yelland (1993) consistently examined gender as a variable in her research. In comparison of mixed and same gender pairs when performing Logo tasks, Yelland found minor differences between all boy and all girl pairs in examining the efficiency of tasks performed. Mixed gender pairs took twice as long to complete the task than did pairs of girls assigned to complete the same task (Yelland, 1994; Yelland, 1995).
Guntermann and Tovar (1987) found several significant results in their study examining the question of differences in social interaction behaviors while using computers. Male groups were observed to display more solidarity than were female or mixed groupings of students. Female group members were observed as much more likely to agree with peers than members of male groups. More questioning behaviors were observed in male groups than in female groups.
Research examining the interaction of cooperative learning and computing consistently uncovers positive effects when that research incorporates gender into research. Dalton, Hannafin and Hooper (1989) found significant gender interactions when examining attitude toward computer use. In Dalton's study, ratings for high-ability students on their attitudes regarding the computer-based instruction were largely unaffected by instructional method. Low-ability females had better attitudes regarding the cooperative computer-based treatment than did low-ability males (Dalton et al., 1989).
The consideration of ability has been a primary factor of investigation in researching computers and cooperative learning. Eraut (1995) reported several ability factors as experimentally significant during 19 different studies. When groups had members of higher academic ability there were positive correlations between ability and the areas of software management, programming experts, subject matter experts, and idea generation.
Groups have often been examined by high and low ability with regard to the experimental task. Researchers have found varying results when examining these factors. In looking at the use of ability as an experimental variable in eighth-grade students, Hooper and Hannafin (1988) found that grouping strategies appeared to have limited influence on high-ability students. Low-ability students grouped heterogeneously appeared to perform at higher levels than did their homogeneously grouped peers. Across several studies, Hooper (1992; 1993) found that instructional efficiency was not diminished when mixed ability members were grouped together.
Experiments (King, 1989) comparing cooperative computer use in groups of high and average ability found high-ability groups used significantly longer statements to describe the experimental task than did average ability groups. Sherman (1994) investigated the use of cued interaction and ability grouping during computer-based instruction (CBI). In his investigations, ability was viewed as a very strong performance predictor when students worked together during a CBI program. Students in lower ability dyads were reported to perform significantly worse on practice items than students in either mixed or higher ability dyads.
Researchers have extensively studied differences between varying group size and individual learners. Amigues and Agostinelli (1992) found that students who worked alone on a computer used the computer more but did not necessarily show increased performance. In contrast, student pairs working together at the computer appeared to be more inclined to work on the solution inherent in the experimental situation and to give their answers more reflection.
Chernick (1990) examined students' performance under three conditions: interdependent, coactive, and individualized computer-based instruction. Significant main effects were found for condition across treatment effects that indicated subjects who worked in cooperative groups performed better than subjects who worked alone.
Eraut (1995) reported on a multi-year program that investigated group composition and group size as one of many research factors. These factors were listed as important for educators to consider when creating groups that will use computers. The first factor noted was the use of rotation, which detailed how often a student used the computer. A second consideration was the level of participation for group members. Group size apparently influenced the amount of dominant behavior by one member of the group. Groups of four seldom experienced one group member as a dominant member. Student marginalization, in which the group rather than one individual limited participation from an individual pupil, was reported to be more common in groups of four and five but occurred rarely in groups of three students. In pairs of subjects, dominance of one person was both more common and less easily remedied. Layout was observed to be a significant factor for groups of four or more students. Four or more students often sat in positions that made it difficult to see the computer screen or printed material. Hooper (1992) found consistently higher scores for students completing experimental tasks in groups when compared to individuals.
Research examining race or ethnicity, as an experimental factor in cooperative computer-based instruction, does not exist within the studies identified. Race and ethnicity have been extensively studied in standard cooperative learning literature.
Training in Cooperative Learning Skills
Johnson, Johnson and Holubec (1993) describe three types of cooperative learning skills. The first, functioning skills, are needed to manage the group effort for task completion and to maintain effective interpersonal relationships. Formulating skills, the second set, provide the mental processes needed to build increased levels of understanding to stimulate the use of higher quality reasoning strategies, and to maximize mastery and retention of the assigned material. Finally, fermenting skills, allow students to participate in academic controversy.
Relatively few researchers have examined the use of training in cooperative learning skills as an experimental variable. Panitz (2001) has several writings promoting the need for training in cooperative and collaborative skills. Hooper, Temiyakarn, and Williams (1993) trained students on the use of cooperative learning techniques but did not measure the effects experimentally. Malouf, Wizer, Pilato and Grogan (1990) investigated the effect of training in cooperative learning techniques as an aid for special education students. Researchers documented an increase in cooperative learning interactions, which was the hypothesized experimental effect.
Repman (1993) compared three conditions in her experiments. Subjects were randomly placed in an unstructured setting, a structured setting, and a structured setting with training. Placing students in cooperative computer-based learning groups resulted in increased achievement in the content area. No differences were discovered regarding measures of critical thinking. Repman proposed that when students are provided with structure and training, then those techniques could be used to enhance instruction.
The central emphasis of this study was to examine if computer-based instruction students, working in pairs and trained in cooperative learning skills, learn differently when compared to individual students completing similar computer-based instruction. The theoretical influences of this study examine an area of theory that contributes to the growth of cooperative learning theory. This study examined the effect of training on computer-based cooperative learning for students aged 18 or above. Researchers have examined the effect of training on fourth-grade and seventh-grade students. Other studies have not examined training on older students.
Relevant theoretical influences from the literature review are presented below.
More than three hundred and fifty studies have examined the effect of training in cooperative learning situations. The ideas of D. Johnson and R. Johnson are the primary theoretical influences for this study. Relatively few other researchers have considered training in cooperative computer based instruction. Rocklin and O’Donnell (1985) considered the effects of training and cooperative learning on university students receiving computer-based instruction. They did not use Johnson and Johnson’s cooperative learning strategies.
Carrier and Sales (1987) examined the differences between individual and cooperative computer-based instruction but did not train students in cooperative learning techniques. Clements and Nastasi (1988) completed a naturalistic comparison of students working together using two different types of computer-based instruction. Students were not trained in cooperative learning strategies but students were trained in Logo and a computer-assisted instructional program. Training in Logo consisted of training in the Logo programming language. The training in computer-assisted instructional programming consisted of instruction regarding five coordinate concepts creating shapes on the computer screen.
Hooper and Temiyakarn (1993) considered the question of training in a total of 175 fourth-grade students who were classified as being of high or average ability and randomly assigned to paired or individual treatments stratified by ability. Students completed training to enhance small-group interaction before completing a computer-based tutorial and a posttest. Following cooperative learning, students demonstrated increased achievement and efficiency as well as better attitudes toward both the computer lesson and grouping. Students completed more practice items and examples in program-control treatments than in learner-control treatments. However, the form of the lesson control did not affect students' achievement or attitudes.
Repman (1993) examined the kinds of elaborated verbal interactions that take place during group processing. In examinations of verbal interactions during group exchanges, most spontaneous student-student verbal interactions appeared to be limited to low level informational exchanges. In this study, the effect of methodology incorporating structure and training (designed to increase the level of elaborated interactions) was investigated. The students participating in the study were a sample of regular and at-risk seventh-grade social studies students engaged in a nine-week program of collaborative computer-based learning. Training led to increased rates of giving explanations and higher self-esteem, while structure (with or without training) resulted in improved content area achievement.
Based on the studies described above, it can be concluded that significant differences exist between individual computer-based learning learners and cooperative computer-based learning. There was no examination of a similar nature that included learners above the age group of 18 years or higher. No study examined the effect of training with cooperative learning and computer-based instruction on college-aged students and adults. Only Hooper and Temiyakarn (1993) considered the cooperative learning strategies elaborated by D. Johnson and R. Johnson (1993) within the area of cooperative computer based instruction.
Researchers investigating the use of cooperative learning as an instructional strategy have examined several variations of this form of learning. Studies examined many individual and group variables: such as gender, ability, training in cooperative learning skills, group size, and other ability factors. With the increased use of computers in classrooms, researchers who found questions surrounding cooperative learning of interest carried these questions into this new research area. Many of these questions have been answered substantially in the area of cooperative computer-based instruction. However, a significant gap remains in the literature regarding the use of training in cooperative computer-based learning. The need for further research into the areas of cooperative learning and computer-based instruction is well established.