Individual conceptual design paper

From VMT

Return to Designing Social Interaction Software

[edit] Conceptual Design Paper

by Stephen Gyarmati

Virtual Math Teams, a project that seeks to create a cohesive online community engaged in the building of mathematics knowledge, is about to embark in a revolutionary new direction in the field of social networks. Instead of creating a typical social network in which the focus is on building casual relationships, Virtual Math Teams is creating a knowledge network, a series of student peers, mentors, and teachers who are linked to one another in a web of communal knowledge generation. In such a community, the focus is not on the relationships between members per say, but on the products that such relationships foster. This revolutionary model brings new responsibilities for the designers of the portal that will give host to the community. VMT will not be best served by laying existing social networking tools on top of whiteboard chat, but by employing the ideas pioneered in existing social networks as an extensible framework upon which new, novel tools can be built. These tools should facilitate the knowledge generation process, aiding in the creation, modification, retrieval, and evaluation of group knowledge artifacts. Ultimately, VMT social networking should be structured so that learning is not an orchestrated byproduct of interaction, but a natural consequence. To meet such an aim, VMT must establish a firm concept of user identity, an effective methodology for locating others to collaborate with, and a system that will link users to solutions.

The basis of any social network is the concept of identity, “who am I within the context of the community?” Elements of identity can include community role, personal information, skills indicators, reputation, and group membership. These various elements are all employed in different ways in social networks. Let us draw from existing implementations. In traditional social networking communities such as the ubiquitously cited Myspace and Facebook, identity is comprised almost exclusively of personal information and popularity. Users are defined by who they claim to be and who they know. Systems such as this are excellent at facilitating acquaintanceship, but would soon fail when placed in the context of an academic community. Academic relationships, such as those between a student and a tutor, are built on a basis of trust. For such trust to exist, objective indicators of ability must gauge a user’s status. Dependence on user entered data to define identity is grossly inappropriate for academic communities. Users may be tempted to mislabel themselves for personal gain or because of some other, more malicious intent.

To solve this problem, purpose driven communities such as Xbox Live use a skills based system that is completely objective, in that it relies on the completion of tasks for the accrual of points. The result is that an individual’s rating is a good indicator of their skill level and expertise in the community. Unfortunately, this implementation, while excellent at pairing users for the purpose of casual gaming, is also inappropriate for an academic community, because it neglects another key factor of trust: disclosure. While, a rating provides an indicator of skill, it fails to let other users know who they are dealing with. Disclosure of personal information can be important for strengthening relationships and transforming peers into acquaintances or even friends who will work together on a regular basis.

Because neither personal details nor skills evaluations alone are sufficient for identity management in an academic community, we must create a new hybrid approach for VMT. Let us detail the various elements of identity within VMT and discuss tools implementations that accommodate them. The first element of identity, user role, is an important consideration for VMT, as within a learning community, this role is extremely fluid. A mentor may become a student, or visa versa, all within the context of the same VMT session. Therefore, rather then assigning structured roles to peers, it is important to place them all on equal footing. System imposed roles are a detriment to interaction, as the learning dynamic is significantly different between peers then between a student and a teacher in the traditional pedagogical model. However, distinctions must be made for system administrators, moderators, or real-world teachers, authority figures whose presence should be made known in the system. These user’s profiles and screen names should be labeled to indicate their position. This presents students with a clear help path, should concerns such as inappropriate user behavior arise.

As for our second element of identity, personal information, it needs to be an important component in profiles, because of the disclosure building trust that was discussed earlier. Nonetheless, the inclusion of personal information, especially that of minors, raises a number of privacy and liability issues. To come to terms with these issues VMT should only allow users who are above the legal age of disclosure to report personally identifiable contact information. Different legal jurisdictions have different policies in regards to the age of disclosure so this too must be taken into account. Apart from the constraints dealing with minors, users of any age should have the opportunity to mask personal information, either by not including it in the first place, or by hiding it from view for select groups. For example, a user could choose to only allow users they have bookmarked as friends to view their contact information. As for the nature of the information itself, users would enter contact information to facilitate out of environment interaction as well as information about grade level, math interests, and outside interests. Social interaction occurs when two individuals have something in common, and the inclusion of some information not relevant to mathematics would aid in producing such interaction.

The aforementioned importance of objective indicators of ability in an academic community necessitates the creation of a system whereby the achievements of users would be tracked within VMT. When students are searching for an individual whom they can trust as an expert, they will typically refer to several criteria in evaluating their peers. Among these are number of problems solved, number of users helped, and length of time involved in the community. VMT needs a system whereby a problem designated as solved could be attributed to one or more users. In addition, users could indicate that they received help from a peer. Ideally, the entry of such data would be mandatory, and take place immediately upon the termination of a whiteboard session. Through the use of methods such as this, use of the system would gradually build up a database of metadata about each user. From this data, various indicators of user quality such as reliability and helpfulness could be extrapolated. These indicators would be listed in a user’s profile along with their personally entered information to serve as a useful tool to peers who are evaluating them. This concept of skills evaluation could be expended to encompass a reputation. Negative actions in the community, such as the disruption of whiteboard sessions or the deliberate misleading of others, would result in penalties to this reputation. Positive actions, such as aiding others or collaborating successfully, would benefit it. Having a low reputation would have real consequences in the community, as peers would be less likely both to seek help from and offer help to such an individual. Thus, by committing antisocial actions contrary to the communities’ best interest, one is effectively ostracizing oneself from VMT.

The final important indicator of identity is membership in groups. Membership in groups is a natural consequence of social behavior as individuals will differentiate themselves and begin to associate with those like themselves. In online communities, establishing the sense of a group is more problematic then in face to face interaction. Whereas face to face interaction offers individuals the chance to define a group through mutual activity or some group created organizational system, online groups are dependent on the constraints imposed by the technology through which they must operate. In online communities, it is not enough to create ad-hoc groups temporarily; instead, special tools must exist for the facilitation and tracking of grouping. In VMT, grouping should be enabled through the use of user created and defined groups. By allowing the users relative freedom in the creation of groups, peers can form a group with membership based on their terms alone, such as geographic proximity or grade level. Open groups, where membership is available to the public, as well as closed groups, where the user creators of the group decide who is admitted, should both exist. Both varieties allow users to perform a type of social bookmarking, in which they join groups that have interests that are pertinent to them. This joining with common individuals can then lead to VMT sessions, in which closely compatible individuals work well together.

Let us discuss the group formation process in more detail. Any user, who wishes to create a group, will enter a page that will allow them to enter various information about the group. The only field that will be completely user defined is the name of the group, the rest of the fields will be limited by the system to certain inputs. This limitation keeps the type of group clear while still allowing a measure of personalization. Fields such as math interests, geographic location, age group, grade level, and common interests will all be limited to prespecified inputs. Therefore, the nature of a group cannot be distorted. This prevents students from being distracted or mislead by groups whose subject matter has been deliberately falsified or distorted. As the creator of a group, a user attains certain administrative rights. They can invite members to the group, remove members from it, and extend administrative rights to others. Thus, a hierarchy of user privilege is established with the group creator as the primary arbiter of power. In addition, should inappropriate activity occur within the group, system moderators do have the ability to shut it down or alter it as they see fit.

The group system would offer unique opportunities for certain individuals on VMT. For examples, for classmates who are already acquainted in traditional classrooms, it would allow them an easy way both to find and coordinate with their classmates. For the group view, VMT needs an online status indicator, showing the number of members that are online and available to partake in whiteboard chat. Additionally, a methodology needs to be provided where one can select one or more currently online users to whiteboard with. An invitation would be extended join that user in a whiteboard, which the invited users could then choose to accept or decline. The invitation window itself would have primary focus on each invited user’s screen; it would pop to the forefront. Joining the session at this point would be as simple as clicking accept. Repeated unnecessary and unwanted invitations could be addressed by allowing users to maintain an ignore list.

Maintaining user profiles as well as a group system is an excellent step towards establishing a community within VMT, but these two tools are not enough by themselves. Browsing through users and groups manually is extremely ineffective and time consuming. Without a search tool, new users would quickly be frustrated and put off from VMT use in the future. Search is important, especially for the novice users, as they seek to establish their identity within the community. Additionally, power users who are well acquainted with the functioning of the system will demand more powerful search parameters that will allow them to tailor their results. Many search criteria should be offered, which can be used either individually or in conjunction. Group search and individual search should be accessible from the same tool. Among the possible search criteria are all the elements of personal information found in the profiles, as well as membership information in groups.

Though having extensive search criteria will be important to power users of VMT, a very complex search tools runs the risk of confusing novices. Also, if a user is facing time constraints and wants to jump into a whiteboard session as fast as possible, extensive search may not be the best option. To cater to these users, VMT should offer quick search functionality. Quick search will provide far fewer options then the advanced search; primarily identifying compatible users and groups based on academic criteria. A user using this tool is not so concerned about pinpoint search accuracy, but about relative matches. In many circumstances an approximate match is sufficient.

Search results, in both the quick and advanced search modes, should place the users and groups that are the best matches to the search criteria, at the top in the search results. Also, online users and groups should be separated from offline users and groups. This accommodates two separate usage scenarios. In one usage scenario, a user is attempting to locate someone to help him within the context of the same VMT session. In this instance, the user only wants to be presented with individuals and groups who can help him immediately. Thus, they need to be presently online. In the second scenario, a user is merely browsing the VMT lobby for users and groups who may be compatible to his interests and may be of use to him in the future. By distinguishing between online and offline status in the search results we enable the first user to gain access to helpful resources immediately, while not depriving the second user of the most comprehensive and pertinent results.

VMT search is a powerful solution to finding resources and is the most appropriate approach in an academic context. Academic users expect accuracy of results and want to be empowered in finding the individuals and resources that are most helpful to them. When faced with the need to get information quickly, academic users do not want to be encumbered by the need to learn the nuances of some matchmaking or pairing system. Also, they will not tolerate inaccurate suggestions by some system that presumes to know what is best for them. Within the context of VMT, manual search is not encumbering; instead, it is empowering. Search is so ubiquitous on modern websites that any somewhat experienced internet user can be expected to understand how it works. The design conventions employed in other search tools have been carried over to VMT. Additionally, through system delimitation of certain search criteria through the use of drop down menus, the user does not need to be familiar with the correct format search queries need to be input in. This measure makes VMT even easier to use for novices.

Thus far VMT social networking has only been seen as a community of users. While the measures discussed so far would turn VMT into an effective social networking portal, they do not embrace collaborative learning in the true sense of the term. Collaborative learning results in the generation of knowledge artifacts, meaningful representations of the knowledge group members have acquired. Within VMT, these knowledge artifacts would most likely be represented textually and graphically. The results of a whiteboard session meet both these criteria; yet, there is no way for such a session to persist once all the users have left. Thus, the knowledge contained within an inactive, completed session is lost. Even if the log from the VMT session were retained and made public, the information contained within is hardly useful to users researching problem solutions. In fact, unless one is actually involved in the session, the whiteboard history can be extremely difficult to decipher. There are many other presentation forms that would be more useful to a user seeking information. VMT needs a way for the knowledge generated by groups to be saved, indexed, and then made searchable.

A few solutions to this problem that have already been discussed, including the introduction of a wiki. However, a wiki presents a new problem: what incentive do users and groups have for entered information into it? Because a wiki is a completely separate tool from the whiteboard, there is no easy method by which to transfer information into it short of manual copying. This cumbersome process calls for a great deal of work and commitment from a user who does not stand to gain anything from doing so. Therefore, the utilization of such a tool is likely to be low unless some reward is attached to the process. Earlier, the important of reputation in a community was established and a system for doing so was delineated. Entering information into the wiki would result in a reputation boost within the community. The wiki creates a system of attribution, whereas solutions are ascribed to certain users. This creates a true knowledge building community, as solutions are archived and their creators acknowledged.

VMT social networking has the potential to set a precedent in online communities. It can demonstrate that not only is collaborative, public, knowledge building possible online, but that primary and secondary students can be a party to such interaction as a supplement to, or component of, their traditional education. If it is successful in its mission VMT may revolutionize online education as we know it. However, the success of VMT depends largely on the structure of its implementation. If VMT strives to address the three aspects of an effective social/collaborative network, it will be successful. User identity must be fostered through profiles, relationships must be built by providing search tools, and user generated solutions must be archived to create a concrete knowledge structure. Without the introduction of any of these critical components, VMT social networking is doomed to failure. Just as in person interaction has fundamental elements necessary for it to exist, so too does online interaction. VMT must take these into account while ensuring that the technology behind its implementation does not hamper the intended learning outcomes of its students.

Works Cited

Goecks, Jeremy, and Elizabeth D. Mynatt. "Leveraging Social Networks for Information Sharing." CSCW (2004).

McDonald, David W. "Recommending Collaboration with Social Networks: a Comparative Evaluation." CHI (2003).

Scardamalia, M, and C Bereiter. "Computer Support for Knowledge-Building Communities." The Journal of the Learning Sciences (1994).

Stahl, Gerry. "Groupware Goes to School."

Stahl, Gerry. "Social Practices of Group Cognition in Virtual Math Teams."

Wellmann, Barry. "A Sociological Perspective on Collaborative Work and Virtual Community." SIGCPR/ SIGMIS (1996).

Return to Designing Social Interaction Software