Analogical Scaffolding in Collaborative Learning Soniya Gadgil (smg58@pitt.edu) Timothy Nokes (nokes@pitt.edu) Learning Research and Development Center University of Pittsburgh Pittsburgh, PA 15260 US Abstract Past research has shown that collaboration can facilitate learning and problem solving (e.g., Azmitia, 1988; Barron, 2000). In the current work, we compared the effects of three collaborative learning conditions: prompts that encourage analogical comparison between examples, prompts that guide sequentially studying single examples, and traditional instruction (practicing problem solving), as students learned to solve physics problems in the domain of rotational kinematics. Preliminary results showed a significant problem type by condition interaction effect. Keywords: analogy; collaborative learning; comparison; problem solving; transfer Introduction Analogical comparison can be a powerful mechanism of learning from examples (e.g., Gentner, Loewenstein, & Thompson, 2003). However, students often have difficulty making spontaneous analogical comparisons (Atkinson, Derry, Renkl, & Wortham, 2000). Recent research by Nokes & VanLehn (2008) has shown that providing prompts to encourage analogical comparison of worked examples improves students’ performance especially on far transfer tests. The current research extends this work by exploring the effect of analogical prompts on collaborative learning. We hypothesize that analogical comparison will scaffold the cognitive processes of explanation and knowledge construction that underlie successful collaborative learning, thereby helping students learn more effectively while collaborating. Much research on collaborative learning has shown that when students learn in dyads, they show better learning gains (at the group level) than working alone (e.g., Azmitia, 1988; Barron, 2000). Much of this research has focused on identifying the conditions that underlie successful collaboration such as the presence of conflict (e.g., Schwartz, Neuman, and Biezuner, 2000), adequate scaffolding of the collaborative interaction (e.g., Rummel and Spada, 2005), and group composition characteristics, such as aptitude, age, gender etc (e.g., Webb, 1982). For example, we know that the presence of cognitive conflict is an important variable underlying successful collaborative learning in particular contexts. Schwartz, Neuman, and Biezuner (2000) showed that when students with misconceptions distinct from each others’ collaborated, they were more likely to learn compared to those with the same misconception, or without a misconception. We also know that scaffolding (or structuring) collaborative interaction is often critical for achieving effective learning gains (see Lin, 2001 for a review). For example, Rummel and Spada (2005) conducted an experiment in which students learned to collaborate by studying an example of collaboration in the presence or absence of a collaboration script. Dyads that received a script showed an advantage in learning over those who received no scaffolding. This is consistent with other results that show that providing scripted problem solving activities (e.g., one participant plays the role of the tutor vs. tutee and then switch) facilitate collaborative learning compared to those who learned individually or in unscripted conditions (McLaren et al., 2007). Hausmann, Chi, and Roy (2004) have identified three mechanisms that are at play during collaborative learning. The first is “other directed explaining” and occurs when one partner explains to the other how to solve a problem. The second is explanation through “co-construction” in which both partners equally share the responsibility of sense-making. Collaborators extend each others’ ideas and jointly work towards a common goal. The third mechanism is “self-explanation” in which one partner is engaged in a knowledge-building activity for his or her own learning. Data from physics problem solving by undergrads showed that all three mechanisms are at play in learning to solve problems collaboratively. However, the former two are more beneficial to both partners while the third is only beneficial to the partner doing the self-explaining. In the current work, we explored whether scaffolding collaborative interaction by the means of providing analogical prompts can help students learn more effectively. We hypothesized that analogical comparison will provide specific scaffolding to encourage other- directed explanation and knowledge co-construction compared to studying individual examples sequentially, thus ensuring that both partners benefit from the collaborative interaction. To test these hypotheses, we conducted an in-vivo classroom experiment in which we had students collaborate under one of three conditions: 1) with comparison prompts (i.e., questions instructing participants to compare two examples), 2) with sequential prompts (i.e., the same questions targeted towards studying individual examples) and 3) without prompts (problem solving and reading expert solutions/ explanations). The results will help us understand whether analogical comparison can be an
[1]
N. Rummel,et al.
Learning to Collaborate: An Instructional Approach to Promoting Collaborative Problem Solving in Computer-Mediated Settings
,
2005
.
[2]
D. Gentner,et al.
Learning and Transfer: A General Role for Analogical Encoding
,
2003
.
[3]
Yair Neuman,et al.
Two Wrongs May Make a Right ... If They Argue Together!
,
2000
.
[4]
D. Dellarosa.
Role of analogical reasoning in the induction of problem categories.
,
1992
.
[5]
K. Holyoak,et al.
Schema induction and analogical transfer
,
1983,
Cognitive Psychology.
[6]
Xiaodong Lin.
Designing metacognitive activities
,
2001
.
[7]
Noreen M. Webb,et al.
Peer interaction and learning in cooperative small groups.
,
1982
.
[8]
Brigid Barron.
Achieving Coordination in Collaborative Problem-Solving Groups
,
2000
.
[9]
Timothy J. Nokes-Malach,et al.
Bridging principles and examples through analogy and explanation
,
2008,
ICLS.
[10]
S. Derry,et al.
Learning from Examples: Instructional Principles from the Worked Examples Research
,
2000
.
[11]
Mark T. Keane.
Constraints on Analogical Mapping: A Comparison of Three Models
,
1994,
Cogn. Sci..
[12]
H. Simon,et al.
Perception in chess
,
1973
.
[13]
Dedre Gentner,et al.
Learning by Analogical Bootstrapping
,
2001
.
[14]
M. Azmitia.
Peer Interaction and Problem Solving: When Are Two Heads Better Than One?.
,
1988
.