Abstract

Group tests are one format for cooperative learning and assessment. However, while groups typically outperform individuals on
group tests, how does cooperation translate into individual learning? This project examines students' responses when allowed
the opportunity to 're-do' quiz answers after completing a group quiz. Thematic analysis is used to identify seven categories for
how students explain their new responses. While some of these categories show a positive impact in helping students to sharpen
definitions and remember class experiences, other categories illustrate students' lack of understanding of course material, inability
to formulate an argument and/or misunderstanding of the assessment context.

Introduction

Cooperative learning is a form of active learning in which students work together toward a common goal in the process of learning (Johnson, Johnson, & Holubec, 1994). This form of learning has been utilized in several contexts such as group discussions, long-term group projects, and group testing. The benefits of cooperative learning and assessment are widely accepted (Brody, & Davidson, 1998; Johnson, Johnson, & Holubec, 1994; Slavin, 1983, 1996). The focus of this project is on group testing and specifically, the impact of group testing on individual learning.

Studies have consistently shown that groups perform higher than individuals during group testing situations. In addition, there are other benefits such as long-term retention of course material (Cortright, & Collins, 2003) and reduction of test-taking anxiety (Russo, & Warren, 1999). However, as Olivera and Staus (2004) note, "results of studies that have examined the impact of groups on individuals are mixed and provide inconclusive evidence of the nature of group effects on individual learning" (p. 442; also see Thompson, & Chapman, 2004). This project explores the impact of cooperative testing on individual learning by examining students' ability to provide verbal elaborations of course material after taking a cooperative test.

I focus upon verbal elaborations because in group testing situations students may identify what the best answer is, but a benchmark for long-term retention and translation of knowledge to other contexts is the ability to explain why a certain answer is the best choice. Verbal elaborations provide one way of having students make visible their learning (Angelo, & Cross, 1993). Hence, the guiding research question for this project is: How does cooperative testing impact students' verbal elaborations of course material?

In the study design, students are allowed an opportunity to 're-do' a maximum of two questions immediately after each quiz with the stipulation that their 're-do' will be graded based on the written explanations they provide for their new answers. I use an analytic approach that is a hybrid of a grounded theory approach and discourse analysis for describing students' explanations. In the following sections, I review relevant literature on cooperative learning, describe the study design in further detail, and describe and analyze my results.

Cooperative Learning and Communication

As described earlier, in its most basic meaning, cooperative learning involves having students work together in the achievement of learning objectives. Johnson, Johnson, and Holubec (1994) more specifically describe the conditions for optimizing cooperative learning. These involve structuring the learning environment such that there is a positive interdependence within the student group, allowing "promotive interaction" (p. 29) in which group members can help each other through interaction, holding individuals accountable for their contributions, fostering positive interpersonal and group communication skills, and allowing the group to reflect upon its own process.

The terms cooperative and collaborative learning are both used to describe conditions in which students work together in their learning. However, cooperation and collaboration differ, and as some would argue, can work against each other in that cooperation focuses on learning 'foundational' knowledge whereas collaboration is geared toward developing new or non-foundational knowledge (Bruffee, 1995). I focus on cooperative learning given that my context of group testing is designed to assess students' understanding of foundational course material, and I reserve other course assignments for allowing students to build new knowledge (Wiggins, & McTighe, 1998).

There are several theories that have been used and tested to explain cooperative learning. For example, Johnson, Johnson, and Holubec (1994) identify social interdependence theory, cognitive-developmental theory, and behavioral learning theory. Slavin (1996) describes four approaches that include motivational, social cohesion, cognitive, and developmental. The perspective assumed in this project aligns most closely to the developmental approach.

The Developmental theory is based on the work of Piaget and Vygotsky and takes a constructivist approach (Johnson, Johnson, & Holubec, 1994). This approach highlights the role of communication in learning in that it is through communication that group members provide and hone explanations and offer criticisms occurs. As Bruffee (1992) explains, thought and 'conversation' are interconnected in that "we think because we can talk, and we think in ways we have learned to talk" (p. 26). In a classroom context, students learn when they are encouraged to articulate and explain their understanding to others and in turn, when they must evaluate and respond to others (Smith & MacGregor, 1992; also see Tompkins & Cheney, 1982). For group examination situations, students must provide explanations for their responses and their peers must critically evaluate those responses in the process of determining the group's response for the test questions.

The work of John Shotter (1993a, 1993b), a social constructionist working from the field of psychology, is closely related to Piaget and Vygotsky's developmental approach. Shotter however, with a focus explicitly on language and communication, adds additional dimensions to consider to understand social interactions. Shotter (1993a) labels his approach a Rhetorical-Responsive Social Constructionist approach. In doing this, he points out that when people communicate it is from within a context in which, in order to demonstrate social accountability and have one's voice be taken 'seriously,' one must be responsive to the given context. Students must, in other words, not only know the material, but also must be able to demonstrate their understanding of the course material in a way that is responsive to their context. However, in order to do this successfully, students must understand not only the course material, but also the rules and demands of the context. Or, as Kaartinen and Kumpulainen (2002) explain, "learning is a participatory process in which the learner gradually becomes an active member in a cultural community by learning its discourse practices, norms and ways of thinking" (p. 191).

Study Design

I conducted this study in Fall 2004 in my Organizational Communication class, an upper-level class consisting of approximately 30 students, most of them majoring in Communication. There were a total of three quizzes administered during the term, consisting of fifteen to twenty multiple-choice questions. This project was a continuation of an earlier project examining group discussion during cooperative exams (Castor, 2004).

The quiz procedures were as follows. First, students take the quiz individually, with a time limit of fifteen minutes. Then, students take the same quiz again in groups of four to five, with a time limit of thirty minutes. In the groups, students decide as a collective their answers, and hand in a group answer sheet. The groups are randomly assigned and students know about a week in advance which students will be in their groups. Also, throughout the semester group work is used so that students become accustomed to small group discussion of course material. Most groups finish the quiz in less than the allotted time. Grades for the quizzes are determined by adding the individual quiz grade with a student's group's quiz grade. In this respect, individual and group accountability are evenly shared.

In Fall 2004, a third portion was added to the quiz. After handing in the group answer sheets, students were allowed to 're-do' two of the answers from their individual portion. Only a sub-set (4-5) of the quiz questions were eligible for re-doing. Usually, these questions called for application of a concept to an example or making distinctions across concepts were selected for re-doing since these would call for students to show a higher level of thinking than questions that focused on recall of definitions. Students had the remainder of the class period, usually five to ten minutes, for re-doing questions. For this portion, students were provided with the following instructions:

In re-doing your answers, identify (a) which question you are re-doing, (b) your original answer, (c) your new answer, and (d) your reason for changing your responses. Your new answer will be graded based on your demonstrated understanding of the course material. If you change a correct answer to an incorrect answer, you will be graded based on the newer response.

For my analysis, I focused on the written explanations the students provided in re-doing their responses. Because group quizzing and re-doing answers after a quiz was a new type of testing for many of the students, I did not collect responses after the first quiz. Also, responses were collected only from students who indicated agreement to participate in this research project by signing an informed consent form. Because of the limiter of using data only after the second and third quizzes and from students who agreed to participate, my total sample size is 24 student responses.

For my method of analysis, I utilized an analytic induction or grounded theory approach (Strauss, & Corbin, 1990). A grounded theory approach is based on allowing categories to 'emerge' from the data rather than begin with categories to apply in coding the data. This process involves the researcher reading and re-reading the available data to create categories that summarize patterns in the data. In this process, I focused on categorizing the type of explanations that students offered in justifying their new responses. My reasoning in selecting a grounded approach to analyzing my data is that because this approach is inductively rather than deductively driven, it privileges understanding students' sensemaking or 'voice' with respect to the course material.

Results

In this section, I first provide a quantitative overview of my results, followed by a qualitative overview. The former provides a summary of the student responses in terms of the number of correct versus incorrect answers. The latter, qualitative overview describes students' explanations in changing their responses.

Quantitative Overview

Table 1 illustrates student responses in terms of correct versus incorrect and credit versus no credit. Out of the 24 responses, 20 consisted of responses in which students changed an incorrect answer to a correct answer, and 4 consisted of responses in which students either changed a correct answer to an incorrect answer or substituted an incorrect answer for another incorrect answer. However, when the persuasiveness of student explanations for the new responses is taken into account, fewer responses earned credit. Specifically, among the 20 correct responses, 12 were awarded credit, and 8 were not. In other words, while 20 answers indicated that students knew what the correct answer was, they did not articulate their reasoning for that correct answer in a way that demonstrated an understanding of the course material. The next section explores the different way students explained their change in answers.

Table 1
Comparison of Student Responses as a Function of Being Correct and Credit Awarded

Qualitative Overview

There were 7 forms that students' explanations took for justifying the new answer. Some individual student responses utilized more than one of these forms. The first four forms described below were not awarded credit. A summary of the 7 forms with a representative example for each is presented in Table 2.

Group Persuasion or Influence. Students most frequently cited (a total of 11 times) that the reason for changing their answer was due to group influence. In these cases, the students did not explain what was persuasive about the group's reasoning or discuss what the student understood regarding the course material. The student basically demonstrated that s/he is susceptible to persuasion.

Personal Error, including Misunderstanding the Question. There were six instances in which the student explained his/her change in response as being due to initially putting down the incorrect answer due to a personal error and asserted that they in fact did understand the concept. However, in some of these instances, the students often failed to present evidence for their understanding or to address the substance or content of the quiz question.

Assertion, no other form provided. In two instances, the students' responses took the form of stating the new answer without attributing any reason for the change or provide any other information to indicate an understanding of the relevant course concept.

Guessed. One student explicitly stated that his/her initial response was just a guess. The student however did not provide any additional explanation as to why the newer response was a more appropriate choice. While guessing is a form of personal error, I separate the categories of personal error from guessing since in the former category, students claim to have actually known the correct answer from the beginning.

While the preceding types of responses did not receive credit, the following three types of responses tended to be awarded credit.

Define the relevant concepts. Eleven student responses basically presented definitions of the course concepts relevant to the questions. These responses often took the form of asserting or stating the definition, rather than referring to an external authority source. Subsequently, sometimes students defined the concepts incorrectly. For example, one student used folk knowledge to define the terms. In other words, s/he defined the terms based on breaking down the key terms using everyday definitions, and did not realize that the definitions were specific to a particular theory.

Class Experience. There were four instances in which students cited a specific class example or experience. As indicated by student responses, this memory was often triggered by the discussion with other group members reminding the student of the context for the concept. In all cases, the discussion focused on a specific example or visual image (i.e., graph on chalkboard, video) described in class.

Re-Analyze Example. Finally, there were three instances in which students would develop a re-analysis of a given example from a quiz question to explain how they developed their new conclusion. In these cases, students made explicit their new reasoning and understanding of a concept in relation to the example provided in a question.

Discussion

There are multiple ways to interpret the student responses. On the plus side, the student responses show some of the different ways that group interaction benefited the understanding of various students. For example, the discussions helped in triggering other students' memories by reminding them of specific examples and explanations for concepts provided from classroom lessons. Also, the cooperative testing situation helped students to view questions and answers from different perspectives through the discussion with their peers. The discussions and opportunity for re-doing responses helped students to define concepts more precisely.

The opportunity to re-do quiz questions also provided a second chance for students who may have made errors in answering the test questions initially due to anxiety and not reading the question and answer options carefully. Other studies have shown that group testing helps in reducing testing anxiety. Therefore, when allowed the chance to re-do question after group discussion, students are enabled to show their understanding in a situation that is less anxiety-filled for them.

On the other hand, there are ways in which the student responses do not bode well in term of showing the efficacy of group testing in relation to student learning. While a majority of the students knew what the 'correct' answer was, they were limited in terms of their ability to explain why. One explanation for this is that the group testing format (which lasted a total of 50 minutes) was too limited an amount of time for students to learn concepts that they had an initially weak understanding of prior to the group test. Also, students may have been motivated to 're-do' their answers as a type of gamble to get more points, as opposed to wanting to demonstrate their newly-honed understanding of course material. Finally, there was a weak interdependence among students for the group testing process. In essence, the group grade was dependent upon the group as a whole indicating the best answer rather than upon every member understanding the material. Therefore, there was no incentive for more knowledgeable group members to make sure that less knowledgeable members understood the material thoroughly.

The student responses also illustrate a mismatch on the part of my assumptions as the instructor and students understanding of the grading context. Specifically, I assumed that students would know how to craft a persuasive argument to show their understanding of the material. Also, I assumed that they clearly understood that their new responses were being graded based on explicitly discussing the course material, as opposed to generally discussing their reasons for changing their answers. The student responses do demonstrate a type of accountability in that the students provided accounts of why, in terms of their decision-making. Their responses, therefore, can be interpreted as a misunderstanding of the context of their responses, as opposed to exclusively being a failure to articulate the course material. In essence, students were not responsive to the context under which they were being graded by not integrating course material into their explanations and formulating an argument in the traditional format of stating a claim and providing evidence and/or reasoning.

Conclusion

Group testing is beneficial for both assessment and learning purposes. While group tests function to assess students' learning, they also provide a context for students to continue to learn from each other. The context of the test provides a motivator for students to discuss, defend, and challenge answers to reach a group goal of achieving a high score on the exam. However, there were some ways in which this testing situation was not truly a cooperative testing situation. For example, students were not truly interdependent. In other words, the group grade was not based on everyone in the group understanding the material, but instead on the group determining the best answer. To do this, not everyone had to be on board. However, while the group had no motivation to ensure that everyone accurately understood the material, individual students who did not understand the material had some motivation for asking for clarification given the re-do opportunity after the group portion.

One implication of the student responses is to illustrate students' misunderstanding of the grading context. In my instructions for the re-do option in asking students to show their understanding of the course material in explaining why they selected their newer response, I assumed that my students would understand how to present an argument, and that they would understand what it means to present reasons. When given the opportunity to earn more points, they failed to demonstrate their understanding of the concepts, and also, correlated with this, failed to demonstrate that they understood how to develop a persuasive argument. Students, through their responses did not seem aware of their rhetorical situation or context, or the need to persuade the instructor and the basis for developing a persuasive argument.

This project shows the various ways in which cooperative testing impacts students' verbal elaborations of course material. Some of the results are promising in terms of emphasizing the importance of cooperation in enhancing individual student learning. Overall, this project demonstrates the positive impact that group testing has on student learning and the benefit of allowing students to show their individual learning resulting from cooperative experiences. In addition, this project demonstrates the need to instruct students in not only the content of the course, but also in how to construct arguments that are responsive to their specific context.

References

Angelo, T. A., & Cross, K. P. (1993). Classroom assessment techniques: A handbook for college teachers (2nd ed.). San Francisco: Jossey-Bass. Author. (2005).

Brody, C. M., & Davidson, N. (Eds.). (1998). Professional development for cooperative learning: Issues and approaches. Albany, NY: State University of New York.

Bruffee, K. (1992). Collaborative learning and the "conversation of mankind." In A. Goodsell, M. Maher, & V. Tinto (eds.), Collaborative Learning: A Sourcebook for Higher Education. University Park, PA: National Center on Postsecondary Teaching, Learning, and Assessment.

Bruffee, K. (1995) Sharing our toys: Cooperative learning versus collaborative learning. Change, 27 [Electronic version]. Retrieved from Academic Search Elite.

Castor, T. R. (2004). Making student thinking visible by examining discussion during group testing. In M. V. Achacoso, & M. D. Svinicki (Eds.), New Directions for Teaching and Learning, Volume 100: Alternative Strategies for Evaluating Student Learning (pp. 95-99). San Francisco, CA: Jossey-Bass.

Cortright, R. N., & Collins, H. L. (2003). Student retention of course content is improved by collaborative-group testing. Advances in Physiology Education, 27, 102-109.

Johnson, D. W., Johnson, R. T., & Holubec, E. J. (1994). The new circles of learning: Cooperation in the classroom and school. Alexandria, VA: Association for Supervision and Curriculum Development.

Kaartinen, S., & Kumpulainen, K. (2002). Collaborative inquiry and the construction of explanations in the learning of science. Learning and Instruction, 12, 189-212.

Olivera, F., & Straus, S. G. (2004). Group-to-individual transfer of learning: Cognitive and social factors. Small group research, 35, 440-465.

Russo, A., & Warren, S. H. (1999). Collaborative test taking. College Teaching, 47 [Electronic version]. Retrieved from Academic Search Elite.

Shotter, J. (1993a). Conversational realities: Constructing life through language. London: Sage.

Shotter, J. (1993b). Cultural politics of everyday life: Social constructionism, rhetoric, and knowing of the third kind. Milton Keynes: Open University Press; and University of Toronto Press.

Slavin, R. E. (1996). Research on cooperative learning and achievement: What we know, what we need to know. Contemporary Educational Psychology, 21, 43-69.

Slavin, R. E. (1983). Cooperative learning. New York: Longman.

Smith, B. L., & MacGregor, J. T. What is Collaborative Learning? In A. Goodsell, M. Maher, & V. Tinto (eds.), Collaborative Learning: A Sourcebook for Higher Education. University Park, PA: National Center on Postsecondary Teaching, Learning, and Assessment.

Strauss, A., & Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage.

Thompson, J. C., & Chapman, E. S. (2004). Effects of cooperative learning on achievement of adult learners in introductory psychology classes. Social behavior and personality, 32, 139-146.

Tompkins, P. K., & Cheney, G. (1983). Account analysis of organizations: Decision making and identification. In L. L. Putnam & M. E. Pacanowsky (Eds.), Communication and organizations: An interpretive approach (pp. 123-146). Beverly Hills: Sage.

Wiggins, G., & McTighe, J. (1998). Understanding by design. Alexandria, VA: Association for Supervision and Curriculum Development.