Although previous studies have acknowledged that helping behavior has many potential benefits, few researches have aimed at understanding which factors would possibly enhance helping behaviors among team members in CSCL environment. Accordingly, this study was intended to identify underlying factors leading learners to collaborate in virtual CSCL settings. A total of 100 undergraduate students enrolled in organizational behavior course participated in this study.
Don’t waste time! Our writers will create an original "CSCL Environment" essay for youCreate order
Participants were divided into 20 work teams and were asked to collaborate with group members to successfully complete their group report on time. This study corroborated that shared identity was a critical preceding antecedent of the engendering of learners’ helping behaviors. According to our empirical results, learners strongly identifying themselves with the group perceived a high sense of group norms and a strong trust in team members, which in turn would result in the delivery of more helping behaviors. Implications for educators and instructors to enhance helping behaviors among team members in CSCL environment are also discussed in this paper.
Shared identity, Group norms, Trust, Helping behavior, CSCL
The potential benefit of computer supported collaborative learning (hereafter CSCL) is to create more opportunities for peer-to-peer interaction than that in a traditional classroom. Learners interact with one another for knowledge exchange by providing help to, or asking for help from, others. Therefore they can learn from each other in many ways, such as recognizing and resolving different viewpoints, and internalizing problem-solving processes and strategies that emerge during group work (King, 1992; Webb, Farivar, & Mastergeorge, 2002; Webb & Palincsar, 1996). Although previous studies have acknowledged that helping behavior has many potential benefits for the immediate work group and the organization (Moorman & Blakely, 1995; Organ, 1990; Podsakoff, MacKenzie, Paine, & Bachrach, 2000; Van Dyne & LePine, 1998), few studies have aimed at understanding which factors would possibly enhance helping behaviors among team members in CSCL environment, thereby the issue stands out as particularly important. In this study, we sought to clarify the effects of trust, group norms and shared identity in order to explain the engendering of the learners’ helping behaviors in CSCL.
First of all, trust is undoubtedly an important variable in influencing interpersonal relations, and would be closely bound up with the positive relationship among people (Tanis & Postmes, 2005). In particular, trust is important to support virtual group because it supports parties to facilitate members to continually cooperate, share information, and believe that their party is “us” (Jarvenpaa & Leidner, 1999). As a result, social exchange relationships cannot develop in the absence of trust (Blau, 1964). Researches (Putnam, 1993; Ring & Van de Ven, 1994) have also demonstrated that when relationships are high in trust, individuals are more willing to wage social exchange in general and, particularly, in cooperative interaction. Thus, from the social exchange perspective, trust inspires a social exchange relationship, making the learners to be more willing to help their partners.
Furthermore, all members of the group are potential help givers. However, before members render substantial assistance, it is required for them to be willing to do so. Being willing to give help depends partly on group norms supporting the behavior of working together and helping others as well as a focus on understanding and learning. Indeed, the impact of the group norms on group members’ behaviors has been a recurring theme in social psychology since its early days (e.g., Asch, 1956). Norms of behaviorally relevant social groups are likely to influence people’s willingness to engage in attitudinally consistent behaviors (Postmes, Haslam, & Swaab, 2005a; Postmes, Spears, Lee, & Novak, 2005b). Therefore, we assume that group norms may be an important variable influencing helping behaviors in CSCL groups because stronger group norms may increase pressures placed on individual group members (Zaccaro, 1984).
Finally, based on the social identity theory (SIT; Tajfel, 1972; Tajfel & Turner, 1986), this study postulates shared identity as a preliminary determinant in explaining one’s perception of trust and norms in a CSCL context. There are two reasons for us to draw on SIT to sustain our suppose. First, social identity is defined as a sense of belonging to a group, which is consistent with the premise of collaborative learning proposed by Rourke (2000). Rourke (2000) remarked that “if students are to offer their tentative ideas to their peers, if they are to critique the ideas of their peers, and if they are to interpret others’ critiques as valuable rather than as personal affronts, certain conditions must exist. Students need to trust each other, feel a sense of warmth and belonging, and feel close to each other before they will engage willfully in collaboration and recognize the collaboration as a valuable experience.” In other words, a sense of group seems to be the first step for collaborative learning. Finally, from psychological perspective, through social identification individuals perceive themselves as psychologically intertwined with the fate of the group. Indeed, empirical evidences have shown that individuals view their group and members more positively than other group members (Hewstone, Rubin, & Willis, 2002; Lewicki & Bunker, 1995; Rimal & Real, 2005; Tanis & Postmes, 2005; Tanis & Postmes, 2007) and are more likely to internalize the normative behaviors as they identified with their group (Haslam, Postmes, & Ellemers, 2003; Postmes et al., 2005a; Postmes et al., 2005b). While works in SIT argue that simply categorizing individuals into a common group is enough to increase their altruism toward the group (Tajfel, 1981; Tyler, 1999), it seems reasonable that making social identity salient is also conducive to increase cooperation. Thus, in this study SIT was referred to as an insightful framework to explain the engendering of learners’ helping behaviors in CSCL settings.
To sum up, the main objective of this paper is to identify underlying motives that lead learners to collaborate, which would eventually help educators tailor appropriate strategies to promote helping behaviors in a CSCL environment. Kreijns, Kirschner and Jochems (2003) propose that social preconditions trust, shared understanding, and accountability will enable effective computer-mediated collaboration. In line with their statement, these social preconditions suggest that achieving and maintaining common ground are essential for collaborative teams to solve problems. Unlike the existing literature focusing primarily on helping skills classification (Webb, et al., 2002; Webb & Palincsar, 1996) or on examining the differences of helping behaviors between face-to-face (hereafter FTF) and computer mediated communication (hereafter CMC) environments (Blair, Thompson, & Wuensch, 2005), this study seeks to investigate certain plausible antecedents that enhance the learners’ helping behaviors in the CSCL environments from the aspect of psychological dimension.
Researches have demonstrated that students derive numerous benefits from working in collaborative groups, for example, by giving and receiving help, sharing knowledge, building on each others’ ideas, recognizing and resolving contradictions between their own and other students’ perspectives (Webb & Palincsar, 1996). From Vygotsky’s (1981) view, cognitive potentially benefit from the helping behaviors embedded in the social interactions, such as giving help and receiving help. Regarding this, three questions are to be raised: (1) are giving help and receiving help equally beneficial? (2) are some types of helping behaviors less helpful? (3) does the medium (FTF or CMC) have an effect on helping behaviors? First, despite the potential benefits, however, researches on social interaction and learning show that giving explanations usually has more beneficial effects than receiving explanations on learning (King, 1992; Palincsar, Anderson, & David, 1993; Webb, et al., 2002). For example, while providing explanations to help others solve problems, learners may generate selfexplanations that help them internalize principles and construct specific inference rules for solving the problem (Webb, et al., 2002). Furthermore, attempting to give explanations may help students monitor their own understanding or help them become aware of misunderstandings or lack of understanding (King, 1992); otherwise, they may falsely assume that they know how to solve the problems.
As to the second question, past researches have distinguished some types of helping behaviors. For example, helping behaviors in organizations are defined as voluntary behaviors so that promoting interpersonal harmony and helping coworkers solve or avoid work-related problems are all referred to as help (Podsakoff et al., 2000). Likewise, Severy and Davis (1971) have distinguished two kinds of helping behaviors: psychological help and task help. Following their generic typology of helping, many other focuses proposed in literature can be reconciled within these two kinds of helping. The former, psychological help, contributes one’s psychological well-being, the primary concern is to relieve others’ distress or making others feel better (Organ, 1990; Pena-Shaff & Nicholls, 2004; Podsakoff et al., 2000; Van Dyne & LePine, 1998). In contrast, task help is concerned with the completion of a recognizable task (Moorman & Blakely, 1995; Organ, 1990; Podsakoff et al., 2000; Smith, Organ, & Near, 1983). In this regard, helping behavior involves helping others with work related problems voluntarily and preventing the occurrence of psychological discomfort. Accordingly, help may come in the form of gaining additional information about the task under consideration. Help may also mean that the student has been encouraged to participate in group task.
The focus of this study is put on both psychological and task help giving. We do not intend to consider the cases of receiving helps, because researches have corroborated that the impact of giving help on learning performance is stronger than receiving help (King, 1992; Palincsar et al., 1993). Furthermore, our focus is on investigating how the learners’ helping behaviors are engendered. Given that the help receiver is a passive agent providing no substantial assistance to the giver at the same time when a helping behavior occurs, it is meaningless for us to take into account the help receiving behavior. Thus, our focus is on the effectiveness of giving help. For giving help to be effective, Webb and Palincsar (1996) provided three conditions that must be met: (1) the help must be relevant to the particular misunderstanding or lack of understanding of the less able peer; (2) the help provided must be at a level of elaboration; (3) the help must be given in close proximity in time to the peer’s request for help. Of the three conditions, the detail level of elaboration seems to play the most significant role in collaborative groups (Bargh & Schul, 1980; Stahl, 1999; Webb & Palincsar, 1996). From a theoretical perspective, elaborated help encourages explainers to clarify and reorganize the material in their own minds to make it understandable to others (Bargh & Schul, 1980). On the contrary, giving non-elaborated help (e.g., only the final answer), delivers fewer benefits, because it may not involve cognitive restructuring or clarifying on the part of the help contributors, thereby diminishing the possibility for help receivers to correct their misconceptions (Webb et al., 2002). However, it is assumed that the task helps that students gave would vary a great deal from detailed explanations to answers only at all.
Finally, the analog between FTF and CMC is far from perfect, and the medium indeed has an impact on helping behavior (Van der Meijden & Veenman, 2005). In their study, Van der Meijden and Veenman invited 84 sixth graders aged 11-12, randomly paired into 42 dyads, to compare peer learning in FTF versus CMC situations. Results of their study showed that while CMC dyads provided less high-level elaboration than the FTF dyads, they provided about twice as many affective utterances as the FTF dyads did when solving mathematic problems. More specifically, they stated that students in a CMC situation tended to be fleeting with short contributions since they devoted considerable time to the relatively superficial aspects of what they typed (i.e., format, spelling) rather than actual contents. Actually, elaborate communication may be particularly difficult to achieve in distributed groups. Straus and Olivera (2000) argued that a number of characteristics of distributed work reduce the level of elaboration; these include the difficulty of initiating contact, the bandwidth of communication media, and reduced opportunities for synchronous work. When using a low bandwidth medium, such as electronic mail or computer conferencing, for instance, the costs of engaging in elaborate communication are high. Along with the line, group members are more likely to share only the final answers rather than the processes they use to reach the solutions because it seems relatively effortless.
Empirical evidences have corroborated that trust is important for successful online interactions (Jarvenpaa & Leidner, 1999; Kanawattanachai & Yoo, 2002). The field of researches concerning trust is broad and encompasses varied approaches, which most commonly focus on trust as a psychological phenomenon. Psychological states are referred to as affective or cognitive process associated with situational contexts and may be influenced by the person’s interaction with situation (Lewicki & Bunker, 1995). McAllister (1995) thus defined interpersonal trust as “the extent to which a person is confident in, and willing to act on the basis of, the words, actions, and decisions of another” (McAllister, 1995, p. 25). He viewed trust as having both cognitive and affective foundations. Cognition-based trust is grounded in individual beliefs about peer reliability and dependability, whereas affect-based trust is ground in reciprocated interpersonal care and concern. In this study, our focus is put on interpersonal level of trust in team members, because it would be influenced by the person’s interaction with situations (Bhattacherjee, 2002; Jarvenpaa & Leidner, 1999; McAllister, 1995).
The willingness to harmoniously collaborate is likely to be contingent on whether their collaboration would put them at risk of being taken advantage of by a partner (Mayer, Davis, & Schoorman, 1995). In this regard, trust refers to a belief about the dependability of the partner, which results in one’s willingness to be vulnerable to others because of an expectation that others will perform actions favorable to his/her interests (Mayer et al., 1995). In situations with high levels of trust, people are more willing to take risks (Dirks, 1999) because they feel safe to do so. Furthermore, trust also refers to an effective means to promote knowledge sharing (Williamson, 1985) and exchange resource (Tsai & Ghoshal, 1998) among team members to accomplish ongoing tasks. These processes mark trust as an important factor in collaborative interaction to solve problems. Therefore, group members who trust each other tend to feel psychologically safe to facilitate helping behaviors because “it alleviates excessive concern about others’ reactions to actions that have the potential for embarrassment or threat” (Edmondson, 1999, p. 355).
Although researches have acknowledged the notion that trust plays a critical role in influencing group cooperation (Jarvenpaa & Leidner, 1999; Williamson, 1985). However, given the absence of a shared work history, and the limited options of communication channels, working in the CSCL settings can possibly be disastrous because it seems harder for group members to gather information and evaluate one another’s behaviors in virtual settings (Kanawattanachai & Yoo, 2002). In contrast, trust is presumed to be easier to generate and sustain when people are spatially clustered because co-location permits greater knowledge of others (Lewicki & Bunker, 1995). More recently, Wilson, Straus and McEvily (2006) documented that trust develops slower in the CSCL settings than it does in the FTF environments. In sum, trust is indeed influenced by the medium (CMC or FTF). Thus it seems worthy for us to shed light on whether the impact caused by the medium on trust would in turn affect helping behavior. In their study, Wilson et al. (2006) showed that trust started lower in the CMC teams; however, these teams eventually develop levels of trust comparable to the FTF teams over time. Based on their findings, trust in CMC teams may not always be lower than that in the FTF teams because of its dynamic nature. Actually, Kanawattanachai and Yoo (2002) also documented consistent result in this regard. In their study, they found that both high- and low-performing teams started with similar levels of trust but high-performing teams were better at developing and maintaining the trust level throughout the project life. While agreeing that trust tends to be influenced by the medium, according to both studies, we supposed that the difference of impacts of trust on helping behavior in CSCL and in FTF would be so slight that may not be of major interest in this research.
To sum up, both studies indicated that trust in CMC groups increases over time and, more specifically, CMC groups eventually reach a level of trust comparable to face-to-face teams at the end of the research period. Given the fact that Wilson et al.’s study was a 3-week period and Kanawattanachai and Yoo was an 8-week period, we suspected that individuals in the CMC group may not take too much time to attain a level of trust that is comparable to the FTF group. According to Jarvenpaa and Leidner (1999), in the absence of FTF interaction, virtual teams are encouraged to build trust swiftly at the very outset. This is also consistent with Meyerson, Weick and Kramer’s (1996) finding, which documented that temporary teams usually developed trust quickly. In fact, a virtual group, like that in CSCL, is formed for a common task with a finite life span. The tight deadlines under which these groups work leave little time for relationship building. The situation of high time pressure enables members to take action and this action will help the team maintain trust and deal with uncertainty, ambiguity, and vulnerability while working on complex interdependent tasks with strangers (Meyerson et al., 1996). Seeing that Wilson et al.’s (2006) 3-week research period is enough for virtual groups to develop levels of trust comparable to FTF groups, this study believed that the impact of trust on helping behavior may not be significantly different between virtual groups and FTF groups, particularly when it is 8-week long in this study and the group task is complex and unstructured.
In conclusion, trust generally refers to the belief that people maintain about the other party’s future behavior. The more learner A believes that learner B will fulfill the latter’s commitments to the relationship, the more learner A will trust learner B. In groups with high level of trust in team members, learners thus feel comfortable in directing their effort toward the group task, because they are not afraid that their partners will take advantage of them. Conversely, when learners distrust team members, they will refrain from freely providing helping behaviors or exchanging knowledge and information. Empirical evidence was supported by Dirks’s (1999) study, which proposed that members in high-trust groups demonstrate more helping behaviors when individuals anticipate that others will not take advantage of their assistance. In this regard, trusting group members is a kind of psychological safety condition which provides individuals with the assurance that knowledge and information will be used for the ongoing task. Based on this perspective, we propose that:
H1: Learner’s trust in team members is positively associated with his/her helping behaviors.
Norms are shared patterns of thought, feeling, and behavior within a group. It refers to a guideline that tells members in a group what they can and cannot do (Hogg & Tindale, 2005; McGrath, 1984). According to Hackman’s (1992) definition, norms typically regulate activity that is important to the group. Further, from theoretical perspective, social comparison theory (Festinger, 1954) suggests that we make assessments about appropriate modes of conduct by comparing ourselves with others in our social midst. When we are unsure about how to behave in a new or unfamiliar situation, we look to the behaviors of others (Tanis & Postmes, 2007). Accordingly, when others are engaged in a behavior, they provide us with social approval cues and we “view a behavior as correct in a given situation to the degree that we see others performing it” (Cialdini, 2001, p. 100).
However, group norms may not always be beneficial, especially when the collective norm is detrimental to individual well-being. From a brief literature review, two kinds of detrimental norms were identified: norms of deviance and norms of distortion. The former, norms of deviance, is mainly from the situation which is characterized by the high prevalence of a behavior. Bystander apathy (Darley & Latane, 1968), for example, is a particularly egregious form of inaction on the part of individuals who perceive prevalence behavior of not helping someone in need as a situation that requires no individual intervention. Regarding to norms of distortion, it is mainly from the distinction between collective and perceived norms. Because collective norms are seldom formally codified or explicitly stated (Cruz, Henningsen, & Williams, 2000), it is likely to be divergent in how people interpret them. For this reason, an aggregation of perceived norms among members of a social system will likely not represent the prevailing collective norms. The false consensus effect (FCE; Ross, Greene, & House, 1977) is one of the cases of the inaccurate assessment of social norms. People use others as sources of information regarding social reality. However, these judgments of social reality may be distorted interpretations caused by the tendency of overestimating support for one’s own beliefs (Ross, Greene, & House, 1977). Therefore, people’s behaviors may also be derived from these potentially erroneous social norms.
Compared to those potentially detrimental norms, one type of normative belief that may become codified as a beneficial norm in work groups is the expectation that members will cooperate with one another (Cialdini, 2001; Wageman, 1995). Because normative behavior is reinforced by society, knowledge that others are behaving in a specific fashion should create pressure on a person to also do so (Festinger, 1954). Specifically, people may reason, “If everyone is doing it, it must be a sensible thing to do” (Cialdini, Reno, & Kallgren, 1990). In this regard, group norms for helping partners may signal that assisting other members is an approved and sanctioned behavior (Cialdini, 2001). In particular, Wageman (1995) indicated that coworkers are likely to rely on norm of reciprocity to govern interactions in the absence of formal or contractual obligations. Therefore group norms in this study specifically represented norms of cooperation because it was the major concern of this study. More specifically, detrimental norms were not to consider since we believed individuals do not blindly copy the acts of others simply because they perceive that others are enacting certain behaviors. In contrast, individuals make assessments about benefits and cost that are likely to result in and they judge the acceptability of the behaviors, leading a high possibility of formation of beneficial norms in problem-solving groups (Rimal & Real, 2005). This is also one of the central tenets of social cognitive theory (Bandura, 1977).
In sum, when a group has a strong group norm for cooperation, members expect each other to engage in information sharing and other behaviors to enhance the completion of task (Cialdini, 2001; Rimal & Real, 2005; Wageman, 1995). Many exchange relationships are driven by a quid pro quo orientation, in which individuals cooperate and help each other to compensate for past help received or in anticipation of help needed in the future (George & Jones, 1997). This shared expectation therefore creates a collective relationship, consequently enhancing learners to feel obligated to help each other and feel responsible for doing so as well, and thereby leading to increased helping and cooperation (George & Jones, 1997). In contrast, groups with weak cooperative norms tend to highlight independence rather than cooperation, diminishing helping behaviors. Accordingly, the following hypothesis is then proposed:
H2: Learner’s group norms is positively associated with his/her helping behaviors.
Social identity was defined as “individual’s knowledge that he belongs to certain social groups together with some emotional and value significant to him of this group membership” (Tajfel, 1972). SIT argues that people classify themselves as belonging to various social categories according to age, gender, socioeconomic status, interests, skills, etc. (Tajfel & Turner, 1986). The underlining assumption of SIT is that the individual feels affinity and desires connection with the referent group, which in turn influences their behaviors (Tyler, 1999). The present study therefore adopted perspective from SIT and assumed shared identity to be the antecedent of group norms and trust in team members because researches have shown that we are influenced not only by the behaviors of others but even more so by behaviors of similar others (Bandura, 1977; Postmes et al., 2005a; Rimal & Real, 2005).
People tend to perceive members of their own groups in relatively positive terms (Hewstone et al., 2002). Indeed, some researches have noted that under conditions of collective identity, other group members are perceived as more trustful (Rimal & Real, 2005; Tanis & Postmes, 2003; Tanis & Postmes, 2005; Tanis & Postmes, 2007). For instance, Rimal and Real (2005) stated that in-group members are typically viewed as being more cooperative, more honest, and more trustworthy than members of other groups since the individual presumes that other in-group members perceive a given trust in similar term and will act in similar fashion. In addition, although trust is expected to have a positive effect on the individuals’ cooperative, voluntary contributions to the group (Dirks, 1999; Edmondson, 1999; Tsai & Ghoshal, 1998), trust alone may not be enough to induce the learners’ discretionary efforts, namely helping behavior in the present study. Whereas psychologists conceptualize trust as a psychological event within the individual, sociologists refer to the conceptualization of trust as a property of collective units (Coleman, 1990), suggesting that trust results from not only an individual’s perception of the characteristics but also qualities of specific groups to be trusted. Drawing on Social Identity model of Deindividuation Effects (SIDE; Reicher, Spears, & Postmes, 1995), empirical evidences also have provided parallel findings in this regard.
It is the prediction from SIDE that when group members’ “individuality” is de-emphasized, their perceptions are more likely to be based on group membership and social identity (Postmes, Spears, & Lea, 1998; Reicher et al., 1995). At first, based on SIDE, Tanis and Postmes (2003) documented that the level of identification with a particular group accentuates the perception of unity of the group, and thereby enhances group members’ positive affection. More specifically, Tanis and Postmes (2005), examined the influence of social identity on perception of trustworthiness and trusting behavior. Their results showed that when the individuals were not identifiable, trusting behavior was based on a shared belief inferred from group membership. To sum up, these empirical evidences suggest that a strong group identity leads individuals to think that their members are more dependable and to display positive evaluations of in-group members (Tanis & Postmes, 2007). Thus, when shared identity is reinforced, fellow group members will be evaluated as more trustful. On the contrary, in the absence of a shared identity, distant group members may not have more faith in other members. Based on the above line of reasoning, the following hypothesis is proposed:
H3: Learners’ shared identity is positively associated with the learners’ trust in team members.
It is assumed that the group norms express important aspects of the group’s identity and that group members are motivated to act in accordance with group norms because it is perceived as the right and proper thing to do (Tajfel & Turner, 1986). Based on this perspective, once identified, individuals are viewed through the lens of the relevant group prototype and are represented in terms of how well they embody the prototype. In fact, this assertion can be derived from several theoretical perspectives, including Social Cognitive Theory (SCT; Bandura, 1977), Referent Informational Influence Theory (RIIT; Turner, Wetherell, & Hogg, 1989) and SIDE. In line with SCT, we are influenced by the actions of the models whom we aspire to become because of the individuals’ outcome expectation. Likewise, RIIT also postulates that when people categorize themselves as a group, their perceptions of the group norm become more extreme. As a result, in the absence of the shared identity, there is no reason to expect group identity to affect the individuals’ behavioral choices. Rather, as the individuals’ identity with members of reference group grows stronger, their compliance with the group behavior will be more observable to other group members. Thus, the correlation between group members’ identification with their referent group and their own intentions to comply normative behaviors should be strengthened by the belief that referent group members are also complying the group norm. For in-group members this means that the emphasis lies on the shared identity, whereas for out-group members such a shared identity is obviously not available.
So far, SIDE theorists have claimed that it is through stronger group identification that depersonalization increases conformity to group norms (Lee, 2004; Postmes et al., 1998; Reicher et al., 1995). For instance, a review by Postmes et al. (1998) found that de-individuation has a reliable effect of increasing adherence to context-specific norms. Furthermore, Lee (2004) documented that when de-individuation is encouraged, it leads to stronger group identification, thereby inducing greater conformity to the majority opinion. Lea (2004) explained it might be that de-individuation has led to a more extreme perception of group norms by obscuring differences among the individual group members. Subsequently, increased discrepancies in opinions might have prompted a shift toward the perceived group norms, to reduce the gap. In line with these theoretical reasoning, it is assumed that when a group identity is salient, a shared norm of cooperation will be likely to be adopted and replicated.
More recently, researches have shifted the focus to the link between process of norm formation and social identity (Postmes et al., 2005a; Postmes et al., 2005b; Postmes et al., 1998;). In their study, Postmes et al. (2005a), for instance, consider that the content of social identity is crucial to determining the form and direction of its social influence over group members. Therefore, they discern between deductive identity and inductive identity. The former means that group members infer social identity as well as norms for individual behavior from the wider social context. As to the inductive identity, the individual contributions of group members serve as input for making individuality and individual distinctiveness a defining feature of the group. Indeed, both the deductive and the inductive identity process appreciated the link between identity and norm formation. Seeing that the group members are all on the basis of collective activity oriented towards a shared goal as it works through a particular task, therefore the existence of a shared identity seems to be an emergent property for the task-oriented group itself (Postmes et al., 2005a). Accordingly, in this study we adopted the notion of deductive identity that norm formation on the basis of understandings is gleaned from group properties. Taken together, numerous compelling evidences from empirical studies and theoretical perspectives indicate that the individuals are more likely to engage themselves in behaviors which are in accordance with the group norms they strongly identify with. It is therefore not surprising for us to assume that the influence of shared identity on group norms increases as the individuals strongly identify with their own group. Accordingly, the following hypothesis is then proposed:
H4: Learners’ shared identity is positively associated with the group norms.
Based on the aforementioned literature review, our research framework was then depicted as 1. Shared identity is the most precedent factor in this study, which was expected to have influences on learners’ perception of trust in team members and group norms. In turn, trust in team members and group norms would take effect on learners’ helping behavior.
A total of 100 undergraduate students enrolled in organizational behaviors course participated in this study. Participants were divided into 20 work teams, each consisting of 5 members. The course required a total of eighteen-week of instruction, lasting from mid-September 2005 to late January 2006. All participants were noticed that the team assignment was required to accomplish through web-based team collaboration. The assignments were highly interdependent because all members were required to participate in the project and all members received the same group score. Among the participants, 24 were females and 76 were males. Participants were aged 23 on average. Because group diversity was not the interest of this study, participants were then randomly grouped. Although it was not controlled, the gender of the group composition had a heterogeneous tendency. Six groups were composed of mixed genders, twelve groups were male only and two groups were female only. ANOVA test was conducted and no significant difference on the research constructs was found among these three heterogeneous groups. Therefore we could rule out the potential effects caused by the group composition of the gender.
In this study, each group expressed its preference for a real-world business cases and then were assigned a particular case, which they presented to the class at the end of the semester. All of the real-world business cases to be solved generally included the development of a strategy based on marketing principles; in some cases they required the design of a product. Although no group took redundant case as its learning task, the natures of all cases were equally unstructured. The requested final product about the task was a case write-up and a twenty-minute presentation, both of which were evaluated by the instructor according to specific written guidelines that were given to students in the course syllabus. During the eight-week web-based team collaboration period, participants were asked to complete a group-based assignment to successfully complete their group report on time.
Throughout the eight-week teamwork, students needed to access the self-developed CSCL web-based system to communicate with other team members. Before participants could successfully login, the system administrator had randomly assigned each participant with a specific ID. The system provided threaded discussion spaces for learners to asynchronously contribute their viewpoints on the case problem, clarify or elaborate their solutions on the case problems. Each group has a corresponding threaded discussion board in the system to enhance a sense of group. All participants could read the course content shared by instructors, whereas the correspondence of every post was visible for only those who belong to the specific group. Moreover, not only could learners type in text-content but also they could attach multimedia files in the post. Indeed, participants could chose electronic whiteboard or electronic recorder embedded in this system to send a multimedia file which contains a drawing or voice to help them clarify, explain, and solidify their ideas. In addition, participants could also use annotation tool to mark up any comment on the course material. Both electronic whiteboard and electronic recorder were developed by JAVA language, whereas the entire CSCL system and the annotation tool were developed by DotNet language. Students were aware that the quality and quantity of their posts would be taken into account in their final grades.
We used five measures to assess corresponding constructs. Subjects were asked to answer each question according to their perceptions on the degree they agree or disagree with the statements (ranging from 1, strongly disagree to 5, strongly agree).
Shared identity was measured by 6 items adopted from Tyler (1999). Trust in team members was measured by 7 items from McAllister’s (1995) and Kanawattanachai and Yoo’s (2002) scales. Although Bhattacherjee (2002) seemed provide a sound scale to measure trust, however, it is not applicable to this study because our major concern was dedicated to interpersonal trust. Indeed, Bhattacherjee believed that individual level trust (such as interpersonal relationships) was quite different from that at firm level. Accordingly, he reconciled many other researches regarding how to measure trust, and thereby developed a scale that was mainly to measure trust in e-commerce contexts, which was particular to measure the customers’ trust on the website. Given the fact that the applicable context and the concern of the scale by Bhattacherjee were quite different, this study therefore adopted items from McAllister’s (1995) and Kanawattanachai and Yoo’s (2002) to measure trust. Group norms were measured by 3 items adapted from Wageman’s (1995) cooperation norm scale. Helping behavior was measured by 7 items from Van Dyne and LePine’s (1998) scale.
Assessment of the research model was conducted using Partial Least Squares (PLS) Version 3.00 Build 1058 to test the model. PLS places minimal restrictions on measurement scales, sample size, and residual distributions (Chin, Marcolin, & Newsted, 2003). Unlike traditional regression analyses and factor analysis, PLS addresses both structural and measurement models at the same time. PLS produces loadings between items and constructs (similar to principal components analysis) and standardized regression coefficients between constructs. PLS was preferred to LISREL for this study since it is prediction-oriented and seeks to maximize the variance explained in constructs. In addition, PLS is primarily intended for predictive analysis in which the explored problems are complex, and theoretical knowledge is scarce.
The model was assessed in two stages: the measurement model assessment and structural model assessment. The measurement model consists of the relationships between the constructs and the indicators used to measure them. The measurement model in PLS is assessed in terms of item loadings, internal consistency, and discriminant validity. For discriminant validity, items should load more strongly on their own construct than on other constructs in the model, and the average variance shared between each construct and its measures should be greater than the variance shared between the construct and other constructs. The structural model and hypotheses are tested by examining the path coefficients (which are standardized betas). In addition to the individual path tests, the explained variance in the dependent constructs is assessed as an indication of the overall predictive strength of the model.
For satisfactory discriminant validity, the square root of the AVE from the construct should be greater than the correlation shared between the construct and other constructs in the model (Fornell & Larcker, 1981). Table 1 listed the correlations among constructs, with the square root of the AVE on the diagonal. The diagonal values exceed the inter-construct correlations, which indicated a satisfactory discriminant validity. Convergent validity refers to that the indicators of a given construct should be highly correlated among themselves. We evaluated the convergent validity of the scale by two criteria suggested by Fornell & Larcker (1981): (1) all indicator factor loadings should be significant and exceed 0.70, and (2) the average variance extracted (AVE) for each construct should exceed 0.5. AVE ranging from 0.54 to 0.72 (Table 1) indicated that all constructs have demonstrated a good convergent validity.
1. Share identity
2. Trust in team members
3. Group norms
4. Helping behavior
aIn Table 1, diagonal elements (shaded) are the square root of the AVE of each construct. Off diagonal elements are the correlations among constructs. For discriminant validity, diagonal elements should be larger than off-diagonal elements.
Reliability was assessed by Cronbach’s alpha coefficients. As shown in Table 2, items for each construct were listed. In addition, Table 2 also showed the mean, standard deviation for each item, as well as the Cronbach’s alpha coefficient and composite reliability of each construct. The Cronbach’s alpha coefficients for each construct were all well above 0.7, which indicated that the scales employed had a good reliability. In addition, most items had a loading of higher than 0.7 on their corresponding constructs, providing evidence of acceptable item convergence on the intended constructs. One exception was the second item of the helping behavior scale, which loading was 0.68. By and large, the overall reliability of the scale was acceptable.
Shared identity (SI)
Composite reliability = 0.92, Cronbach’s alpha = 0.89
I am pleased to be a member of the group.
I can count on the group to help me when I need help
The group is willing to help me solve problems
I am proud to tell others that I am part of the group.
I would recommend to close friends that they join the group.
I am proud to think of myself as a member of the group.
Trust in team members (T)
Composite reliability = 0.90, Cronbach’s alpha = 0.88
I can rely on other teammates not to make my job more difficult by careless work
Most of my teammates can be relied upon to do as they say they will do
If I shared my problems with my team. I know (s)he would respond constructively and caringly
I would feel a sense of loss if one of us was transferred and we could no longer work together
Most of my teammates approach his/her job with professionalism and dedication
I can talk freely to my team about difficulties I am having at work and know that my team will want to listen
I see no reason to doubt my teammates’ competence and preparation for the job
Group norms (GN)
Composite reliability = 0.88, Cronbach’s alpha = 0.80
In my group, we expect everyone to assist one another in order to benefit the group.
My group’s norm is to help one another with our assigned group tasks.
In my group, we think that everyone should volunteer to do things for the group.
Helping Behaviors (HB)
Composite reliability = 0.89, Cronbach’s alpha = 0.86
We regularly take time to out ways to improve our group’s work processes.
We would seek new information that leads us to make important changes.
I have presented information or have discussions with other members to improve our work.
I have helped other group members with their work responsibilities.
I have assisted other group members in their work for the benefit of the group.
I would get involved to benefit the group.
I would volunteer to do things for the group.
As shown in 2, the effects of trust in team members (β= .254, p<0.05) and group norms (β= .511, p<0.05) on helping behavior were significant. These results showed that trust in team members and group norms were significant factors in enhancing helping behaviors in CSCL environment. Hence Hypothesis 1 and Hypothesis 2 were supported. Hypothesis 3 posited a positive link between shared identity and trust in team members. The path was positive and significant (β= .742, p<0.001), supporting the contention that shared identity increases the level of trust in team members. Hypothesis 4 supposing that shared identity enhances group norms was also supported. We found that shared identity had significantly positive effect on group norms (β= .657, p<0.001). This result indicated that learners having higher level of shared identity perceived a higher level of group norms, thus supporting Hypothesis 4
Given that Hypothesis 3 and Hypothesis 4 were supported, we were also interested in assessing whether shared identity would directly affect helping behavior. To evaluate this relation, we followed the four-step procedure outlined by Baron and Kenny (1986) to test the mediating effects of group norms and trust in team members. It was first established that the antecedent (i.e., shared identity) was correlated with both (a) the outcome (i.e., helping behaviors) and (b) the proposed mediator (i.e., trust in team members and group norms) and then (c) the proposed mediator was correlated with the outcome. Finally, the last step was to determine (d) whether in a regression using both the antecedent and the mediator as independent variables to predict the outcome, the unique effect of the antecedent was significantly reduced (indicating partial mediation) or no longer reached significance (indicating full mediation). As can be seen in Table 3, the models examining group norms (Model 3a) and trust in team members (Model 3b) as potential mediators indicated that these factors only partially mediated the effects of shared identity. However, Model 4 that included group norms and trust in team members in the same step revealed that, in combination, both variables fully mediated the effect of shared identity. Taken together, these models suggested that one pathway through which learners’ shared identity shaped their helping behaviors was by influencing their perception of group norms and trust in team members.
In addition, we also followed Baron and Kenny’s (1986) suggestion and used moderated regression analysis to further test the moderating effects of group norms and trust in team members. In step 1, we entered all the variables shared identity, group norms and group trust. In step 2, we entered the interaction terms for both shared identity and group norms, and shared identity and group trust. As can be seen in Table 3, the interaction effect for shared identity and group norms was not significant (Model 5a). Likewise, the interaction effect for shared identity and trust in team members was not significant too (Model 5b).
Trust in team member
Shared identity ×
Shared identity ×
Trust in team member
*: p < 0.05, **: p < 0.01, ***: p < 0.001
To sum up, the model explained fifty-seven percent of the variance in helping behavior, fiftyone percent of the variance in trust in team members, and forty-three percent of the variance in group norms. The results suggested that shared identity, group trust and group norms played important roles in shaping helping behaviors. In overall, the proposed model fitted the data pretty well and the empirical evidence provided a strong support for us to believe that trust in team members and group norms fully mediated the relationship between shared identity and helping behaviors.
In order to further understand the predictive power of the model in explaining learners’ actual helping behaviors, we referred to learners’ actual helping behaviors as psychological helping and task helping (Severy & Davis, 1971). We conducted a post hoc analysis and chose content analysis as the main methodology to analyze learners’ online discussion. The initial category as shown in Table 4 was used as a guideline to create the coding scheme. These nine categories were mainly based on Stahl’s (1999) criteria to characterize the quality of knowledge-building and the work by Webb et al. (2002) and Pena-Shaff and Nicholls (2004). Specifically, we distinguished between high-level task help and low-level task help to characterize the contribution of the learners because learners’ elaboration was identified as high- and low-level (Van der Meijden & Veenman, 2005). In this regard, low-level task help included two categories, namely: answers only and general explanation. High-level task help encompassed four categories, namely: organization, analysis, articulation and brainstorming. Psychological help in this study was based on the work developed by Pena-Shaff and Nicholls (2004). Psychological help was mainly a statement not related to formal content of subject matter or a part of statement that encouraged group members to participate in group task. Psychological help included three categories, namely: acknowledgement, empathy and greeting.
Data obtained from the discussion board were coded and categorized by evaluating each message with reference to the type of information presented. Two researchers (one doctoral student and one graduated student) independently examined each message and created a coding scheme based on their understanding. The percentage agreement was found to be 63%.
Type of helping
Low-level task help
Offering correct answers without showing how or why.
Offering comments or data that are already available without further developing an idea in the message thread.
High-level task help
Organizing of existing thoughts or perspectives
compares or contrasts previously articulated perspectives or derives new understandings from existing data
Explaining of complex or difficult concepts
Introduces new ideas concerned with the topic or task and provides a perspective not previously considered
Acknowledging other participants’ contributions and ideas
Sharing of feelings with other participants’ comments
Social comments not related to the task at hand
Note: adapted and modified from Stahl (1999), Webb et al. (1998) and Pena-Shaff and Nicholls (2004).
After identifying high-level task help, low-level task help and psychological help, we had learners’ actual helping behaviors. They were treated as dependent constructs to examine the predictive power of the model in explaining learners’ actual helping behaviors. As shown in 3, perceived helping behaviors measured by instrument were learners’ subjective perception of helping, whereas high-level task help, low-level task help and psychological help categorized by the raters were objective and actual helping. The coefficients of the effects of trust in team members and group norms on low-level task help were .330 and .237, respectively. Overall, only low-level task help and perceived help were significant, indicating that although the predictive power on perceived help was corroborated, this model only partially predicted learners’ actual helping behaviors.
*: p < 0.05, **: p < 0.01, ***: p < 0.001
The purpose of this study was to identify underlying motives for learners to collaborate in CSCL settings, which would eventually help educators tailor appropriate strategies to promote helping behaviors in CSCL environments. We expected shared identity to be the most precedent factor in influencing learners’ trust in team members and group norms. In turn, trust in team members and group norms enhanced learners’ helping behaviors.
The empirical evidences revealed that shared identity had indirect effects on enhancing helping behaviors through trust in team members and group norms. The results were consistent with the previous studies (Hewstone et al., 2002; Rimal & Real, 2005; Tanis & Postmes, 2005), which confirmed that individuals strongly identifying with the groups are more likely to believe that their members are reliable and then reduce the level of uncertainty while collaborate virtually. Our results were also in line with the findings provided by Kanawattanachai and Yoo (2002), which indicated that high performing teams would maintain a high level of trust among team members throughout the project. Based on our findings, we further asserted that those high performing teams maintaining a high level of trust in team members may due to a high level of shared identity. A high level of shared identity encouraged members to trust their partners, and thereby enhanced their willingness to provide help. By giving and receiving helps, each team member shared knowledge and internalized problem-solving processes that emerged during group work (Webb & Palincsar, 1996), thereby increasing team performance. Therefore, we concluded that those teams whose members with a strong shared identity would outperform than other teams. Future research may examine this assertion to see if it is supported.
Our findings also confirmed that as a learner recognizes more himself/herself as a member of the group, he/she would perceive stronger group norms. This finding was consistent with the past researches (Cialdini, 2001; Rimal & Real, 2005; Postmes et al., 2005a, 2005b). Accordingly, learners with higher level of shared identity were more likely to engage in helping behaviors needed to enhance the relationship with other group members. In sum, shared identity had impacts on both trust in team members and group norms, which were elements of relational dimension of social capital (Putnam, 1993). Along with the line, individuals in CSCL setting developed and exchanged relational dimension of social capital by recognizing the self as an embodiment of the in-group prototype. Although Putnam (1993) suggested that social capital facilitates cooperation for mutual benefit, this study, more specifically, suggested that learners in CSCL required a shared identity before they could benefit from relational dimension of social capital. With stronger shared identity, learners internalized group norms as prototypes that governed their behavior, and were more willing to believe their partners are dependable and would not take advantage of their assistance. Apparently, learners were used to employee relational dimension of social capital within the group that they strongly identified with while encountering problems. Thus, an interesting question is whether shared identity could also be used to explain the other two dimensions of social capital, such as cognitive and structural dimension, in CSCL environment. There is a need for further research to test and validate these relationships.
Consistent with the past research (Dirks, 1999), this study found trust had positively associated with helping behaviors. These assistances underlay in such a way that learners believed they could not succeed unless they coordinated their efforts with the efforts of their group mates. More specifically, learners monitored their contributions in light of their partners’ contribution and made adjustments accordingly. Therefore, reciprocal helping happened in that you help me, I’ll help you. In other words, the possibility of mutually beneficial exchange became more probable as long as learners believed other members were dependable to return favors.
Another factor that had direct effect on enhancing helping behaviors was group norms. The finding was consistent with the past researches (Cialdini, 2001; Rimal & Real, 2005), which confirmed the positive effects of norms on behavior. However, the role of norms in attitude-behavior relations may not always be consistent. From social identity perspective, norms must be prescriptive rather than descriptive to influence attitude-behavior consistency. Prescriptive norms provides a template for how one should think, feel and behave, whereas descriptive norms means most people who are important to one think that one should perform the behavior. Based on this perspective, the nature of the norms of cooperation in this study seems to be a prescriptive one. Therefore, we suspect that prescriptive norms would be derived from deductive identity because deductive identity considers that group members infer norms for individual behavior from the wider social context. In this regard, prescriptive norms seems share the common basis. As a result, group members with a high level of group norms perceived a clearer template for what should be done, which also regulated their behavior toward others.
Following Baron and Kenny’s (1986) guidelines to examine the mediating and moderating effects of group norms and trust in team members on helping behavior, this study seems provide a slightly inconsistent finding while comparing to Tanis and Postmes’s (2005) study. In sum, their results showed that individuals’ trusting behavior (transferring money) was based on expectations of reciprocity (they would be rewarded), not on perceived trustworthiness. However, our results documented that helping behavior was based on both group norms and trust in team members. As a result, there might be a potential moderator variable which leads to unexpectedly inconsistent findings (Baron & Kenny, 1986). We thereby suppose that whether trust has an impact on group members’ certain behaviors may be contingent on the explicitness of a shared goal in the group. While lacking of a shared goal for each group in Tanis and Postmes’s study, there was an explicit objective for each group to accomplish in this study, and therefore group members were highly interdependent. Under the frame of a shared mission needed to finish, learners were more likely to conduct a reciprocal helping, because they knew that not helping other members in need, the group task may not be finished successfully. Therefore, as long as learners have adopted the collective mission and have trusted in their partners, the possibility of mutually beneficial exchange behaviors becomes more probable.
Finally, a post hoc analysis was conducted to test the predictive power of the research model on learners’ actual helping behaviors. The result indicated that it seems there are other factor that influences learners’ high-level task help and psychological help. As to high-level task help, the result was consistent with Van der Meijden and Veenman’s (2005) findings that CMC learners usually employee less high-level elaboration. Straus and Olivera (2000) argued that the costs of engaging in detailed elaboration are high is the reason why learners reduce the level of elaboration. Regarding to psychological help, it seems that learners are used to share this kind of help no matter how strong their perception of group norms and trust in team members. Compared to the high cost of engaging high-level elaboration, providing acknowledgement, empathy and greeting seems has relatively low cost. As a result, most learners were more likely to provide this kind of affective support. Based on this contention, learners’ perception of cost to conduct a certain behavior seems a potential preceding determinant other than shared identity. Future research attempting to account for the engendering of learners’ helping behaviors should take this into consideration.
Collaborative group learning is a group of learners who interact through various communication technologies to accomplish its common goal. Studies have shown that cooperative learning in traditional classroom promotes learners’ achievement within problem solving as well as productivity (Kreijns et al., 2003). More specifically, this study sheds some light on this issue by showing helping behaviors is the product from how strongly learners identify with the group they belong to. In conclusion, the research presented provides insight into three findings:
Unlike existing literature focus primarily on helping skills classification or on examining the differences of helping behaviors between FTF and CMC settings, this study identified plausible antecedents that enhance learners’ helping behavior in CSCL environment from aspect of psychological dimension.
This study confirmed that the notion that shared identity could account for the engendering of learners’ helping behavior. According to our results, learners strongly identified themselves with their group perceived stronger group norms and trust in team members, consequently leading more helping behaviors.
This study concluded that relational dimension of social capital, i.e., trust in team members and group norms, plays an important role in directly shaping learners’ helping behaviors in CSCL settings. In addition, based on the results, other dimensions of social capital, explicitness of a shared goal and perception of cost to conduct a certain behavior were derived and were suspected to have effects on explaining the learners’ helping behavior too.
The research presented in this paper corroborated the importance of share identity for helping behaviors. According to such findings, there are several ways for educators and instructors to enhance helping behaviors among team members in CSCL environment.
Firstly, promoting trust in team members would enhance learners’ sense of dependability. A high level of trust can reduce the negative effects usually present in non-collaborative groups such as the social loafing (Latane, Williams, & Harkins, 1979) and sucker effect (Kerr, 1983). In sum, both of which suggest that groups refuse to further support noncontributing members and therefore reduce their individual efforts. With a high level of trust in team members, learners seem have no reason to doubt their members’ competence and preparation for the task, and thereby diminishing social loafing and sucker effect in the group. Consequently, educators and instructors should refer to learners’ perception of trust in team members as a crucial control, coordination mechanism, and an essential ingredient towards developing successful helping behaviors in CSCL. This would be helpful for learners to share problems with other members to obtain constructive and caring response.
Secondly, group norms is another factor that promotes helping behaviors among team members. One of the weaknesses of cooperative learning is the “diffusion of responsibility” (Slavin, 1983). Some learners may take a greater role in leading the group and do a major share of the work, while others do less work or none at all. It is clear that assigning students to groups does not necessarily mean that they will work collaboratively. Group norms have characteristic of reciprocity that benefits individuals and particularly the whole group. Consequently, for educators in CSCL, this may imply that more attention needs to be paid to develop strategies that help the reciprocal norm formation. Rather than attempting to enhance all types of interaction, instructors should consider focusing on certain types of interaction which is helpful in establishing norms of cooperation to enhance learners’ thinking of their party is “us”.
Finally, CSCL providers and educators should actively seek ways to enhance learners’ shared identity, because it has significant effects on learners’ perception of trust and norms to help others. In the presence of a strong shared identity, members of in-groups are subsequently evaluated more favorably than those considered part of out-group (Hogg & Tindale, 2005). However, this may lead us to a pitfall, namely taking for granted that a shared identity will automatically occur just because CSCL allows online collaborative group learning. Thus, future research considering shared identity as an antecedent would further enhance our understanding of helping behavior in CSCL environments. In addition, prior research has suggested a significant relationship between individual motive and social identity (Stets & Burke, 2000). As long as the shared identity is activated, the ongoing process of trusting in team members and group norms formation will link learners to be more willing to help each others. Therefore, future research can examine whether intrinsic motives and extrinsic motives have any influence on the formation of share identity.
Results of this study represented a first step in understanding the engendering of learners’ helping behavior in CSCL setting, but several limitations are needed to consider. These limitations were discussed in this section so that they could be addressed and improved upon in future research studies. One of the limitations of this study is that most participants in this study have been acquainted with one another. Familiarity contributes to shared understanding, thereby may have provided learners with shared identity before their collaboration online. Although we do not know whether there is difference between shared identity generated collocately and shared identity generated distantly, it is obvious that learners’ perception from the social context may change over time when they are getting familiar with one another. Therefore, it is reasonable to suspect that group member familiarity may have influenced learners’ perception of shared identity. In particular, familiarity may have led individuals to serve as central inputs, making individuality and individual distinctiveness a defining feature of the group (Postemes et al., 2005a). As a result, learners may have developed specific norms before participating in this research. Accordingly, future studies should reduce the potential impact of group member familiarity by administering the treatment at different geographic locations (e.g., different campuses).
Another limitation of the research is from the gap between students’ perceived helping behavior and their group members’ perceived effectiveness and quality of the behavior. According to Webb et al. (2002), an elaboration to be an effective help must be relevant, detailed and prompt. In this study we classified different level of detail of elaborations posted by the participants, however, whether a help is relevant, detailed and prompt enough, eventually, is determined by the help receivers. Lacking of the help receivers’ perception of quality of each help given by their group members, we are unable to point out the conditions under which one feels that one have helped one’s group members, but that others do not agree. However, having this kind of gap between the help givers and receivers is not astonishing. In fact, group norms are essentially a kind of belief rather than a comprehensive guideline of course of action. Therefore learners who strongly perceive norms of cooperation may have more willingness to help others, however, being willing to continue giving help until the learner feel helpful may be quite another.
Moreover, the limited choices of communication technology in this study (thread-style discussion) may not be realistic when other forms of technology such as video conferencing, instant messaging, and other groupware tools are widely available, accessible and affordable. Furthermore, the data was only gathered from college students, further research is necessary to verify the generalizability of our findings. In addition, common method bias exists due to same respondents provide information on both independent and dependent variables. Finally, group trust in this study is an overall general trust although many researchers considered trust to be multi-dimensions. Future research may decompose trust into several dimensions and examine their respective effects.
We will send an essay sample to you in 2 Hours. If you need help faster you can always use our custom writing service.Get help with my paper