تأثیرات اجتماعی بر رشد کودکان از تفکر ارتباطی در طول بحث در گروه های کوچک

This microgenetic study strived to understand instantaneous peer influences on the moment-by-moment and session-by-session development of relational thinking within and across dialogic small-group discussions using an approach called Collaborative Reasoning. An analysis encompassing 32,511 turns for speaking during 176 discussions indicated that peer support and refutation influenced the development of relational thinking within (micro-level) and across (macro-level) discussions, and was mediated by friendship and peer status. Support was mainly mediated by friends and children with high status. Observing reciprocated friends' supportive talk encouraged students to generate confirmational relational thinking in the next turn for speaking. Refutation was mainly mediated by children with high status. Quiet students generated less refutation. The study documents the proximal effects of peer status and friendship on the social and cognitive dynamics of collaborative discussions.

Relational thinking is the ability to perceive, construct, and manipulate relations between concepts to form coherent schemas (Holyoak, 2012). Understanding the mechanisms by which this ability develops is crucial to understanding cognitive development, as relations are the building blocks of all kinds of knowledge (Dumas, Alexander, & Grossnickle, 2013). Previous studies of relational thinking highlight the importance of individual cognitive factors (e.g., Halford et al, 2002 and Gentner, Rattermann, 1991). So far, however, social influences on the development of relational thinking have received little attention. The major goal of this study was to capture instantaneous social effects on moment-by-moment cognitive development during and across collaborative small-group discussions. The general working hypothesis was that large-scale changes in thinking depend upon many small steps that are made possible by recurrent patterns of productive dialogic interaction. The study modeled the time course of support and refutation that students offered each other during the give-and-take of a socially-supportive, cognitively-engaging small-group discussion approach called Collaborative Reasoning (Chinn, Anderson, & Waggoner, 2001), examined whether these elements of discussion were associated with the micro-development of relational thinking, and explored whether effects were mediated by peer relationships. We theorize that Collaborative Reasoning (CR) discussions provide a context where students can socialize their relational thinking through a dynamic co-construction and co-evaluation process (Anderson et al., 2001); the emphasis on social support in CR reinforces positive peer relationships, which contribute to socially harmonious and cognitively invigorating interaction. Studies of cognitive development typically evaluate students' growth in terms of change in pre- and post-intervention assessments, which does not shed light on how and when students' cognitive skills progress, and especially what types of interaction bring about change. Little is known, for example, about whether friends are more willing to support or oppose each other, and whether popular students contribute to or hinder group processes. The study employs the microgenetic method ( Siegler, 2006). An essential feature of the method is that the density of observations is high relative to the rate of change in the phenomenon under investigation. As Siegler (2005) explains, “Learning tends to follow irregular paths involving regressions as well as progress, short-lived transitional approaches, inconsistent patterns of generalization, and other complexities. Because of this complexity, the only way to determine how children learn is to follow them closely while they are learning” (p. 770). The present study involved additional layers of complexity, beyond those in most previous microgenetic studies (e.g., Kuhn et al, 2008 and Siegler, Svetina, 2002). Instead of the behavior of individual students, the collective action of groups of students was tracked. Instead of discrete trials under the control of the experimenter, the students freely controlled their own behavior. 1.1. Relational thinking in collaborative discussions Relational thinking involves the ability to appreciate how things are connected (e.g., predator–prey relationship, kinship relationships), the ability to identify common structures among things with distinct surface features (Chi & VanLehn, 2012), and the ability to manipulate these relations to form systematic concepts or schemas (Hummel & Holyoak, 2005). The ability to recognize and manipulate complex relations enables students to perform many higher-order thinking functions, such as drawing inferences between premises and conclusions to reach logical coherence during reading, generating analogies in argumentation, identifying abstract mathematical principles, associating theory and evidence in scientific discovery. Relational thinking is fundamentally important to knowledge transfer and conceptual change (Holyoak, 2012 and Schwartz et al, 2011). Many individual cognitive factors have been found to contribute to developmental change (Gentner, Rattermann, 1991, Goswami, 1991, Halford et al, 2002, Loewenstein, Gentner, 2005, Richland et al, 2010 and Richland et al, 2006). Less is known about moment-by-moment development during a social process, how interpersonal factors influence the process, and how micro-level development contributes to macro-level development over days, weeks, or months. Dialogic interaction is argumentative discourse in which participants all have rights to formulate arguments to support their own viewpoints and probe others to better understand or refute opposing viewpoints (Reznitskaya et al., 2009). A supporting argument embeds relational thinking when it involves reasons or evidence that justifies a claim. Similarly, a refutational argument is relational when counter-reasons or counter-evidence are provided. To generate a supporting or refutational argument, the learners must have some understanding of how ideas can be connected. Means for connecting ideas can be secured by appropriating relational thinking strategies encountered in dialogic talk. We hypothesize that when students observe peers engage in a relational thinking strategy judged to have explanatory power or persuasive force, they are likely to emulate the strategy. We assume that students who frequently and successfully generate relational thinking serve as models for those who seldom do or do so less effectively. Subsequently, we suppose that as less-skilled students attempt relational thinking strategies, they are often provided with support by more competent peers. Exposure to various points of view prompts students to compare and contrast perspectives and identify gaps in understanding, which in turn is assumed to advance students' relational thinking. The current study explored the moment-by-moment time course of the emergence of relational thinking in order to evaluate the hypothesized social process. Our theory is that instantaneous social events that embody modeling, support, and refutation are the precursors of growth in relational thinking in the long run. 1.2. Peer relationships in small group discussions Small group discussion can be conceptualized as two interweaving networks: an argumentation network in which individuals are expressed as nodes and the connections between individuals' expressed ideas are denoted as ties; and, a social network in which individuals are connected by ties of friendship or status in the classroom social network. Dynamic cognitive and affective ties, involving disagreement, support, power, or friendship, represent multiple dimensions of co-regulation or co-ordination among group members (Vauras, Salonen, & Kinnunen, 2008). Although the idea of a dual-space framework is not new (e.g., Barron, 2003 and Olivera, Straus, 2004), the majority of collaborative learning research focuses on one dimension or the other (Van den Bossche, Gijselaers, Segers, & Kirschner, 2006). There is not yet a consensus as to how the cognitive and social facets of collaboration interlace (Ladd, Kochenderfer-Ladd, Visconti, & Ettekal, 2012). Particularly, the role of peer relationships in collaborative learning contexts is still unclear (Tolmie et al., 2010). Previous studies suggest that progress in cognitive development depends upon positive peer relationships. Students prefer to interact with or seek help from peers who are popular, have more good ideas, or share many characteristics with them. These selection processes thus may influence when and how learning takes place. For example, Azmitia and Montgomery (1993) found that friends are more likely than acquaintances to evaluate, justify, and criticize each other's ideas, which in turn improves their cognitive performance. According to this line of research, peer relationships are a determinant of cognitive development during peer collaboration. However, Tolmie et al. (2010) argued that optimal peer relationships at best establish “sufficient minima to permit further growth as part of productive activity” (p. 188). Research suggests that social structures can alter individual behavior. A meta-analysis by Roseth, Johnson, and Johnson (2008) concluded that social contexts featuring a cooperative goal structure, as opposed to competitive or individual goal structures, affords greater opportunities for individuals to cultivate positive peer relationships and improves academic achievement. Other studies indicate that an egalitarian social norm can promote positive social behavior, whereas a rigid social dominance hierarchy may foster peer rejection or aggressive behavior (e.g., Chang, 2004 and Cohen, Lotan, 1995). Cohen and her colleagues developed an approach in which children were taught that intellectual abilities are multidimensional, such that everyone is gifted in some way, and in which the teacher made a point of publicly recognizing the intellectual contributions of low-status children. Similarly, Boaler (2008) taught students to be respectful of each other and to fairly consider various points of view. These interventions successfully fostered students' interpersonal accountability and positive interdependence and promoted learning, suggesting that micro-level social learning is determined partly by macro-level social structures (Vauras et al., 2008). The current study therefore assumes that positive peer relationships can have sustaining facilitative effects on cognitive development provided positive social norms such as respect and support are embraced. Consistent with previous research (Faris & Felmlee, 2011), we assume that students who have higher status in the classroom are more likely to take a leadership role by conforming to the collaborative social norms, modeling desired cognitive and social actions, and supporting classmates who conform to norms. Based on these assumptions, this study examined the influence of two important facets of peer relationships – friendship and status in the social network, within an approach to small-group discussion featuring relational equity and critical thinking. Peer status is a key interpersonal factor that may affect learning. Cohen (1994) defined status as “an agreed-on rank order where it is generally felt to be better to be high than low rank” (p. 23). A person's rank is determined by social evaluation, and does not exist in one single form. According to Lease, Musgrove, and Axelrod (2002), peer status is a multi-dimensional construct. Several dimensions of status have been studied in the social development literature, including but not limited to social preference based on liked-most and liked-least nominations ( Newcomb & Bagwell, 1995), social dominance or the relative ability to control material and social resources ( Hawley, 1999), perceived popularity ( Cillessen & Rose, 2005), and social network centrality ( Faris, Felmlee, 2011 and Farmer, Rodkin, 1996). In the context of classroom learning, peer status is often associated with academic standing. For example, Cohen and Lotan (1995) measured students' status with ‘math and science competence’ nominations as well as ‘best friend’ nominations. A student's status influences teachers' and classmates' expectations about competence (Berger, Cohen, & Zelditch, 1972). High-status students are expected to have a greater influence on group processes than low-status students. Research has suggested that differential expectations due to status strongly affect students' opportunities to participate in small groups and therefore also affect students' learning gains (Cohen & Lotan, 1995). The current study considered three dimensions of peer status. The first dimension was network centrality, which is defined in this study as an individual's relative position in a friendship network. Individuals who are located in a more centered position in a friendship network tend to be more influential than students in a peripheral location ( Faris & Felmlee, 2011). Since socially centered or high-status students have more social connections, they may play an important role in cognitive socialization. However, research has shown that socially centered students may have negative instead of positive influences on each others. For instance, Ellis, Dumas, Mahdy, and Wolfe (2012) found that higher-centrality adolescents interacted with their peers in more dominant and controlling ways than lower-centrality students. One possible reason for the inconsistent findings is that group norms moderate the influence of socially centered students. Socially centered students may be more sensitive to norms and, once adapted to positive group norms, may play an important role in scaffolding relational thinking during dialogic interaction. Thus, network centrality can amplify either positive or negative effects (see Wang & Eccles, 2012), but we assume that only recurrent positive interaction leads to better social and cognitive outcomes. The second dimension of status was perceived cognitive status. Academic reputation among peers is a significant predictor of self-concept, effort, and average grades after controlling for initial levels of these variables (Gest, Rulison, Davidson, & Welsh, 2008). Chiu and Khoo (2005) found that whether a small group could solve a mathematic problem successfully was influenced by contributions of students perceived to be good in mathematics. We hypothesized that students with a higher perceived cognitive status might generate more relational thinking during small-group discussions and have a greater impact on the cognitive processes during dialogic interaction. The third dimension of status was peer reputation for quietness. Research shows that quiet students tend to be judged as lower in credibility, attraction, and success than more talkative counterparts (Coplan et al, 2011 and Daley et al, 1977). Quiet students are often ignored by teachers and peers (Evans, 2001). In small-group settings where talk is the principal medium of knowledge exchange, students who are quiet may experience difficulty interacting with others because they are perceived to lack ability or knowledge. On the other hand, quiet students may be active thinkers during classroom discussions even though they do not talk much (Townsend, 1998). The current study therefore hypothesized that the influence of quietness was contingent on social context. A small-group activity in which quiet students are respected and invited to participate might facilitate these students' cognitive engagement during discussions. Friendship is a second important aspect of peer relationships. Early friendship research focusing on the relations between friendship and cognitive outcomes suggested that children learn better when working with friends. Friends can provide a sense of relatedness, belonging, and emotional support, which create a safe and secure context for children to exchange information in a joint problem solving space ( Azmitia & Montgomery, 1993), and to be more open to disagreement than nonfriends ( Zajac & Hartup, 1997). However, opposite findings were reviewed by Newcomb and Bagwell (1995) and Zajac and Hartup (1997), who commented that friends were more concerned with resolving disagreements and tend to produce more agreement during collaboration than groups composed of non-friends. Balkundi and Harrison (2006) suggested that a high-density friendship network may “bind individual team members into mutual consensus and lack of disagreement” (p. 61) because friends do not want to hurt the relationships that they have devoted time and effort to maintain. Friends may therefore face a dilemma over whether they should strive for better performance or focus on maintaining high affiliation and agreement among group members. Not until recent years have researchers investigated the mechanisms by which friendship affects learning. Berndt, Laychat, and Park (1990) found that friend dyads' achievement motivation became more similar after having a discussion. Decisions among friends shifted toward an alternative action involving greater achievement motivation when discussions were harmonious and featured reasons that supported the alternative action with greater achievement motivation. Altermatt and Broady (2009) conducted one of the few studies that examined friendship effects by direct observation. A sequential analysis showed that when learners were having trouble solving a difficult puzzle, their help-seeking depended on how their performance was evaluated by friends. This microgenetic study suggested that children not only emulate and internalize friends' behavior but also are influenced by friends' evaluation of their performance. The affective bonds between friends and the inclination to maintain friendships may lead children to align their social goals or values with their friends' (Barry & Wentzel, 2006). Following these process-oriented friendship studies, we hypothesized that children might be biased toward ideas generated by their friends, but additionally that the cognitive effort devoted to supporting their friends may lead to deeper information processing as indicated by the emergence of relational thinking in talk.