Social Categorization: In-Groups and Out-Groups

The Social Psychology of Online Collaborative Learning 11

of IGI Global is prohibited.

As noted, team members in computer-supported collaborative learning en-

vironments have a tendency to automatically assume that distant-site team

members are social loafers. They can also make many other unpleasant as-

sumptions about their distant-site team members. Online collaborative teams

seem to automatically divide themselves into “us versus them” (e.g., Harasty,

1997; Stephan, 1985) resulting in stereotyping and potential bias. Sometimes

the “us versus them” bias involves one site pitted against another. In other

cases, some team members bond, while others do not, and those that bond

become the “us,” whereas those who do not become the “them.”

This tendency is explained in social psychological terms through the use of

in-groups and out-groups. An in-group is a group to whom you, as a person,

belong, and anyone else who is perceived as belonging to that group. In-

group members have positive views of each other, and give each member

preferential treatment. An out-group consists of anyone who does not belong

to your group. Out-groups are viewed more negatively, and receive inferior

treatment in comparison to that of in-group members. In-group members are

perceived as being heterogeneous, and as having positive qualities, referred

to as in-group differentiation (e.g., Lambert, 1995; Linville & Fischer, 1993).

Out-group members are perceived as being “all the same,” homogeneous, and

as having more negative qualities. This is referred to as the homogeneity bias

(e.g., Linville, Fischer, & Salovey, 1989). These concepts are used to explain

hostility between social groups (e.g., Republicans versus Democrats, gays

versus straights, whites versus blacks). Relatedly, this bias creates problems

with teams becoming cohesive across distant sites, as a result of team mem-

bers perceiving students from their site (or those they bonded with) as ”our

team,” and automatically seeing students from the distant site (or those they

have not bonded with) as not part of “our team.” In CORAL, for example,

one site is located in a rural area, and the other is located in a suburban east

coast area. We often nd that students from the rural area view the students

at the east coast area as rude and pushy, whereas the east coast students view

the rural area students as slackers because they are slower-moving. Again,

however, there are methods to reduce this social categorization and associated

hostilities (e.g., Gaertner, Mann, Murrell, & Dovidio, 1989).

12 Ashcraft & Treadwell

of IGI Global is prohibited.

Recommendations

Increase intergroup contact. One proven method for reducing social catego-

rization is to increase intergroup contact, referred to as the contact hypothesis

(e.g., Pettigrew, 1997). It is vital that all team members communicate exten-

sively, in order to reduce cross-site conict and stereotyping. Perkins and

Giordano (2004), as well as many others (e.g., Birenbaum, 2004; Scheer et

al., 2004), also note the importance of encouraging communication, especially

in distance learning. Extensive communication permits team members to see

similarities with others, fostering both synchronous and asynchronous com-

munication with cross-site team members, hence reducing homogeneity bias.

In CORAL, for example, we encourage teams to meet in chat rooms once or

twice a week, in addition to meeting via video conference during class time,

and utilizing discussion boards for asynchronous communication. It should

be noted, however, that in order for increased intergroup contact to have

the desired effect, the overall interactions must be neutral to positive. If the

majority of cross-site interactions are unpleasant and negative, the hostility

between groups will remain or increase.

Introduce superordinate goals. A second method for reducing social cat-

egorization is the introduction of superordinate goals (Sherif, 1958). As

mentioned in the previous sections, the introduction of a task that can only

be met through the efforts of all team members can signicantly reduce the

hostility between in-groups and out-groups, and increase team cohesion.

By working together, team members begin to know each other as unique

individuals, thereby eliminating some of the bias and hostility that is often

found in multiple-site learning communities.

Recategorization. Another consideration for cross-site in-groups and out-

groups is recategorization (e.g., Gaertner et al., 1989). Recategorization

involves changing the boundaries of the in-group and out-group. While some

teams cannot overcome the initial cross-site “us versus them” division, most

teams can. But, when teams are able to overcome initial social categoriza-

tion, other types of in-groups and out-groups can emerge. For example, at

the beginning of the semester, we nd cross-site social categorization to be

very common, but as the semester progresses, team members are able to

make connections with cross-site team members, who then become part of

the in-group. Occasionally, the entire team becomes one in-group, a very

favorable occurrence for collaborative learning. But, when only some team

members bond across sites, the complexion of the team takes on a different

The Social Psychology of Online Collaborative Learning 13

of IGI Global is prohibited.

look. In-groups emerge and consist of both same-site and cross-site team

members, and the same for out-groups. We nd that students who remain in

the out-group tend to have work habits that are not conducive to team efforts

and do not feel favorable to working as a team member. They are resistant

to team work and try to give the impression that they are members of the

team, but it is only an attempt to please authority gures (e.g., professors).

They tend to be social loafers, or communicate less with the team, or are

unpleasant to work with, regardless of which site they are located. Although

teams can continue to work somewhat effectively with minimal contribution

from these out-group members, it is obviously to the teams’ benet to be

inclusive. Thus, we encourage groups of students to form whole teams that

consist of all team members, but if they cannot—say, for personality conict

reasons—we instruct teams to continue to give those out-group members

opportunities to work and become part of the in-group. However, teams are

also coached to have a back-up plan if the work of the out-group member is

not up to par with other team members, or not completed at all.

If recategorization does not occur naturally within the cross-site team, then

we encourage it by asking students to work in pairs across two sites on indi-

vidual sections of assignments. This allows cross-site team members to get to

know each other as individuals, note their strengths, and see them complete

work and convey this information to other team members at their site. In

other cases where collaborative classes are purely Web-based and students

bond over technology-assisted communication, asking in-group students to

pair with out-group students should also have the desired effect.

Cognitive Distortions

We’ve mentioned that students often dislike group work because the learner

had earlier negative group experiences where they felt responsible for com-

pleting all—or most—of the assignment adequately, and without the aid

of group members. In some cases, team members believe that others will

complete the assignment, and as a result, students fail to contribute. There-

fore, other team members have to assume responsibility, and do complete

the assignment alone. In other cases, students behave this way due to a lack

of condence in fellow classmates’ ability to complete assignments to their

standards. They believe that their academic skills are superior to those of

14 Ashcraft & Treadwell

of IGI Global is prohibited.

their teammates, and that their teammates’ quality of work will negatively

affect their grade (e.g., Felder & Brent, 1994; NISE, 1997). In this case, other

team members are willing to contribute to the completion of the assignment,

but are not allowed to do so.

This is an example of a cognitive distortion called the self-serving bias; in

other words, the tendency to attribute positive outcomes to internal causes,

and negative outcomes to external causes (e.g., Brown & Rogers, 1991;

Miller & Ross, 1975). Relatedly, the ultimate attribution error is a tendency

to make more attering attributions about members of one’s own group than

about members of another group (Hewstone, Bond, & Wan, 1983). These

attributions are detrimental to the formation of a collaborative learning com-

munity, and reective of in-group/out-group biases. As a result, these types

of individuals often think the team succeeded only because of their efforts,

or their in-group’s efforts, in completing a task. These individuals often

attribute negative outcomes to out-group members. This perception, while

occasionally true, is more often a cognitive distortion, an illusion manifested

by these individuals, and calling attention to this concept may reduce some

of the ill will that can develop in early stages of collaborative learning.

Relatedly, this type of cognitive distortion is especially common in certain

high-achieving students. While some excellent students are quite adept at

online collaboration, others are painfully unprepared for the experience. They

often feel as though they are the only team members capable of completing

adequate work, and are often dissatised with the work others produce. They

therefore complete whole assignments alone, but then complain that they

have completed all the work, and that no other team members are working.

Other team members, in turn, can feel insulted by this lack of trust in their

abilities, along with being referred to as social loafers by individuals who

consider themselves better-quality students. In reality, this is not effective

learning behavior for the individual, nor for their team. While this sense of

responsibility and independence has been rewarded in other educational set-

tings, it is contradictory to the purpose of online learning communities, and

generally to collaborative learning.

Recommendations

Teach trust and mentoring. Because the reward structure in computer-sup-

ported collaborative learning environments is so different than that in tradi-

tional learning settings, these high-achieving students can feel frustrated and

The Social Psychology of Online Collaborative Learning 15

of IGI Global is prohibited.

betrayed. What they are lacking is a sense of trust in working with others.

Thus, taking time to help them trust their teammates is usually productive.

For example, in CORAL, we often ask these types of students to take a

chance, reduce their workload, and give other team members an opportunity

to contribute. If they can force themselves to back off, they are often pleas-

antly surprised by the amount—and quality—of work their teammates can

contribute. In addition, they need to be shown that it is their responsibility

to help their teammates learn course material. Students such as these must

be taught to be less independent and more concerned about the well-being

of their team members instead of their own individual sense of well-being.

Furthermore, they need to realize how their behavior is actually hindering

team development and the learning of other team members.

Intellectualize. It is helpful, with this type of student, to intellectualize

these experiences by labeling them as the self-serving bias or the ultimate

attribution error, as a strategy to reduce feelings of discomfort that can be

associated when challenging the appropriateness of their behavior. In effect,

it is suggested that teams engage in metacognition (i.e., observe their own

behaviors, apply labels to those behaviors, and determine whether they are

appropriate for team development). If the behaviors are not helpful to team

development, then their task is to develop solutions for those inappropri-

ate behaviors. Not only do these metacognitive exercises help students to

intellectualize and understand unpleasant online experiences, but they also

contribute to developing a life-long learning process (Birenbaum, 2004;

Kitsantas & Dabbagh, 2004).

Stages of Group Development

Students (and faculty) are sometimes reluctant to utilize collaborative learning,

because they are uncomfortable with, and unprepared for, team conict and

conict resolution (e.g., Felder & Brent, 1994). However, it is also useful to

understand that long-term groups tend to pass through a number of stages,

one of which is characterized by disagreement, ranging from mild to more

extensive.

Tuckman and Jensen (1977) suggested that groups go through ve stages

of development, from their inception through their adjournment: forming,

storming, norming, performing, and adjourning. Each has unique charac-