SocialWire: Social Software for Informal Learning

M. E. Sousa-Vieira, J. C. López-Ardao, M. Rodríguez-Pérez and M. Fernández-Veiga

Department of Telematics Engineering, University of Vigo, Vigo, Spain

Abstract. Despite the fact that online social networks (OSN) are widely rec-

ognized as a powerful vector for adding a new social dimension to the learning

management systems (LMS), OSNs do not fully integrate the speciﬁc features of

the learning process yet. In this paper, we report the design and implementation

of a software platform that leverages on the basic capabilities of an OSN and ex-

tends the functionality toward its use in education. It does so by the embedding

of a variety of collaborative activities of informal nature, jointly with tailored as-

sessment procedures and an associated reputation system for the users. The latter

is essential in order to foster the students’ engagement and improve their learning

performance. This software system, though still work in progress, has been in use

for two years in different subject areas. The outcomes and the feedback provided

both by teachers and students are encouraging.

1 Introduction

It is with some delay compared to how other ﬁelds in the society and the economy

have adapted, but it ﬁnally seems that the profound changes driven by the information

technologies are reshaping the education system, even higher education. The move is

universal, only its pace varies, and it is motivated from both the supply and the demand

sides. The new cohorts of university freshmen comprise students who consume and pro-

duce digital contents as habits of their lives. Similarly, the learning materials they are

exposed to are increasingly based on audiovisual contents, or on the intensive use of

computer simulations for studying more easily many complex systems in every imagin-

able academic discipline. In the end, the Internet has substantially lowered any barriers,

either physical, economic or social, against the open access to quality education. The

popularity of many online LMSs (e.g., Moodle), used as tools for organizing courses

and students, and the recent surge of MoOCs (Massive online Open Courses) supported

by the best universities in the world are two signs of this trend (see the Coursera or

edX initiatives [1, 2]). The majority of the software tools developed to aid in these on-

line learning processes exploited the nearly limitless possibilities that web technologies

offer to link distributed documents, transfer them efﬁciently to the users, and display

the information through a graphical interface consistent across devices. However, those

tools assume that the underlying learning process remains unaltered, namely it is still a

student-based effort, rigorously constrained by the instructors’ decisions. Consequently,

several works have pointed out that the efﬁcacy with such approach depends mostly on

the (prior) existence of strong willingness to learn in the students.

So, if the social dimension is not incorporated into the learning platforms, it is likely

E. Sousa-Vieira M., C. López Ardao J., Rodríguez-Pérez M. and Fernández-Veiga M..

SocialWire: Social Software for Informal Learning.

DOI: 10.5220/0004602601330145

In Proceedings of the 2nd International Workshop on Interaction Design in Educational Environments (IDEE-2013), pages 133-145

ISBN: 978-989-8565-65-5

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

that online education will be relegated to reach students in remote areas, to be narrow-

focused toward highly specialized ﬁelds or to be useful only to self-motivated persons.

Pertaining to a cohesive group (in the real world or in a virtual environment) not only

reinforces the experience of learning, but also turns out to be a key factor for the activa-

tion of informal learning processes [3–5]. These happen when the goal is to assimilate

the non-explicit knowledge that spontaneously ﬂows among the group members, re-

lated to the skills and command of a given subject. For instance, the typical approaches

in problem-solving, the de facto professional practices, and so on. In summary, an ef-

fective social network speeds up the task of learning those contextual skills, by means

of behavioral emulation or by mutual recommendation, where recommendation applies

equally to people or to items of content [6].

OSNs over Internet form a powerful environment to add a desirable social dimen-

sion to more traditional online learning systems. But, since the design of typical LMSs

does not have a clean interface with OSNs, integrating the capabilities of a OSN into a

learning platform is not entirely trivial. In this work, we describe the architecture and

development of a software system [7] that extends the basic functionality in ELGG [8]

—a popular, open source software engine for building OSNs— and enables the use of

informal learning processes mentioned above. This enriched hybrid of OSN-LMS may

also be useful for enhancing the generic skills of the students, for instance their criticism

or leadership skills [9]. Nonetheless, that is not the primary beneﬁt, in our view. Rather,

we envision the system as an enabler to design and execute informal learning activities.

These informal activities are of wide scope, but share some common features: i) they

are speciﬁc activities, not necessarily formal, like answer a question, seek more infor-

mation about a topic, participate in a challenge, etc.; ii) informal learning activities are

measurable, in that every task of this kind must produce a deliverable (a piece of text,

a report, a list of information resources, a computer program,...); iii) informal learning

activities are open, in other words, there is possibly more than one valid response or

deliverable object to them. In contrast, one should keep in mind that formal contents

already have well-deﬁned assessment procedures in current LMSs.

Preliminary tests of SocialWire in a real environment are being carried in our col-

lege, among undergraduate students in subjects of the Communications Engineering

curriculum. Nonetheless, aside from our personal focus on engineering education, the

design of the tool is neutral and generic and it might be useful in any other ﬁeld. As a

previous step, we have presented it to a group of e-learning experts to get their opin-

ion. The feedback obtained encouraged us to enhance the initial functionality presented

in [10] and develop new modules.

In the rest of the paper we describe the main components of the software system.

The functionality of the ELGG OSN engine is brieﬂy reviewed in Section 2, for com-

pleteness. Section 3 gives the details of a number of software modules that, as plugins,

have been developed and tested to fulﬁll the requirements of social interactions in our

platform. In Section 4 we report some results obtained with the use of the system.

Lastly, Section 5 summarizes the main conclusions gathered after using the system.

134

2 System Overview

SocialWire, our OSN-enriched learning platform, is based on ELGG, a widely used open

source OSN engine.

The core of elgg uses a uniﬁed data model to manage the different objects that

can exist in the ecosystem of an OSN (users, messages, posts, etc.). The basic and

more general attributes of any object are enclosed in the class ElggEntity, but the

extensibility and ﬂexibility of elgg relies on three other classes for the purposes of

linking objects, tuning their attributes and deﬁne actions to execute on object instances.

Speciﬁcally, class ElggRelationship can connect virtually any two objects. The

connection entails a relationship between the two, it could be friendship, sharing in-

formation, membership to a common group or something alike. Particularization of the

attributes that describe an object is done via objects in Class ElggMetadata. The

semantics of the new attributes is not constrained, so most of the ﬂexibility in elgg

comes precisely from this type of object. Finally, the actions that are to be executed

on a given object can be deﬁned with class ElggAnnotation, giving the users the

privilege of modifying the objects’ behavior and add more actions.

Thus, by combining objects in those classes, elgg can be extended easily into an

online social environment oriented toward educational applications and informal learn-

ing. We did so, developing several software modules to support the outlined function-

ality. Each module is actually a plugin, a companion software piece to the elgg core

that handles the new classes an objects devised to convert the original platform in a

OSN-based learning product. The interaction between those plugins and the core takes

place through message-passing of events, actions (hooks and callbacks) and views. Ac-

tions and events are code scripts invoked as a result of the users’ actions in the interface.

Together, events drivethe internal state of the OSN, understood here as a ﬁnite state ma-

chine, whereas actions are the responses generated by a chain of events. In turn, views

deﬁne the way objects are exactly rendered or presented to the users. Thus, views are

essential to manipulate objects in a organized, systematic way.

A shortcoming in the elgg engine is its lack of support for subgroups, i.e., com-

munities of students undertaking the same tasks, assignments or with shared interests.

Thus, we extended elgg to allow the creation of arbitrary subgroups within a parent

group. Membership to a subgroup entails access to the same capabilities as in the larger

groups, only with a restricted audience, duration, etc.

The description of the several modules (plugins) that comprise SocialWire is the

matter of the next Section.

3 The Software Modules

3.1 Questions and Answers

This plugin’s functionality is fairly obvious: some user pose questions and other users

in his/her group may attempt a response. Answers are curated either by teachers or by

other group members.

The workﬂow of question creation is part in the design of this plugin. To be spe-

ciﬁc, both students or instructors can be authors of a question in a group. In the former

135

case, the question ought to be approved by an instructor, who also assigns points to it

according to his/her judgement. When the question appears open, a period to provide

answers starts. That time period can be deﬁnite or indeﬁnite, depending on parameters

used at the creation time. And the answers can be either individual or collective. While

open, any answer uploaded can be edited or corrected by its authors, unless the answer

has already received an evaluation by the instructors. In this case, the particular answer

is closed (it remains visible) and becomes inmutable.

Points awarded to a question or to an answer serve two purposes simultaneously.

First, to classify questions and answer, that is, using the points as a measure of rele-

vance/importance of that item. Indeed, the plugin can display a thread of answers not

by chronological order, but by relevance or ranking order. Second, points awarded to a

question or an answer are meant to be distributed among its authors, using rules deﬁned

elsewhere. Consequently, this plugin interacts with a generic ranking plugin, one that

collects the points assigned to tasks in any other plugin, sorts the scores by user (or

subgroup) and, ultimately, builds a reputation for every user in the group.

Visibility of questions and answers is under direct control of the instructors, as a

primary mechanism to restrict inaccurate, vague or misleading threads of discussion.

As an illustration, Figure 1 shows the appearance of the questions and answers plugin.

3.2 Challenges

Challenges are in SocialWire synonymous of strategic games. More clearly, a challenge

is in our view any proposed activity asking for hints, ideas, full solutions or suggestions

to a complex problem, one with a nontrivial or non-unique answer. Without loss of gen-

erality, when we identify challenges with strategic games, we are taking the assumption

that challenges are always cooperation (and not competitive) games, i.e., the goal of

a challenge is to achieve collectively an approximate solution to a difﬁcult question.

Hence, there are not incentives for confusing, misguide or take proﬁt from other par-

ticipants.

As before, a challenge can receive answers during a deﬁnite or an indeﬁnite

time period, as the authors decided. But, being intrinsically open, challenges are graded

differently than ordinary questions. Here, the answers do not receive points while the

challenge is open to new contributions. Instead, after the challenge has been closed,

two possibilities exist. One option is that the instructors distribute points between the

different answers, according to their quality, completeness or any sound criterion. Al-

ternatively, the answers (possibly curated of ﬁltrated by the instructors) can enter a poll,

whereby the students themselves vote for the closest to solve the challenge. Then, points

are awarded proportionally to the votes collected by each answer. The context and topic

of the challenge determines which of the two options suits better.

Clearly, this plugin interacts closely with the ranking plugin, too. See Figure 2 for a

screenshot of an example challenge.

Pure competition, albeit not in the most general form, can be achieved with the questions

and answers plugin, simply closing the question as soon as the ﬁrst correct answer has been

provided.

136

Fig.1. Questions Listing.

3.3 Quizzes and Exams

More traditional assessment and grading of the students is naturally included in Social-

Wire. This is exactly the task of the quizzes and exams plugin. Both quizzes and exams

are, basically, sets of questions that must be answered in limited timed in the most exact

and concise way. Maybe the main novelty in SocialWire comes from the fact that a quiz

or exam can be proposed to the entire group (or a subgroup of it). Grading of exams

and quizzes is automatically computed by the plugin, but the authors can distinguish

137

Fig.2. View of the challenge module.

138

between grading a qualiﬁcation (with a weight, if this exam is part of a sequence, for

example), and rewarding the answers with points, like in other plugins. Though assign-

ing points to an exam may seem useless or counterintuitive —what incentive would

a student have to undertake it?—, one should keep in mind, ﬁrst, that self-assessment

allows a student to compare his/her progress with that of other classmates. Relative per-

formance is usually a very strong feedback for the students, especially in the lowest and

highest extremes of the range. And secondly, SocialWire is designed to build a ranking

or reputation upon the users’ activity. Solving exams is a direct way to measure global

comprehension and particular abilities, if desired.

Regarding the interactions with other plugins, part of the design in this case is also

common to the questions and answers plugin, as expected. Due to the dual nature of

assessment (grades or points), the outcomes of this plugin are submitted to the ranking

plugin or to the gradebook plugin, as needed. For ease of use, the composition of an

exam relies on the existence of question pools (by topic and level of difﬁculty), which

are described next. To ﬁnish, Figure 3 displays the appearance of this plugin in a typical

example case.

Fig.3. View of the quizzes and exams module.

3.4 Question Pools

It is a tedious task to compose exams, which tend to be similar across years, once and

over. The plugin question pools allows the authors to store past questions (and their

answers) in a database, so that could be reused later for other purposes. A complex

database (e.g., relational) is not needed at all, hence the name “question pool”. Ques-

tions in the pool are internally organized by means of tags describing the question top-

ics and the difﬁculty level. Using these tags, a simple search engine can easily ﬁnd and

classify subsets of questions in a given pool.

Only the instructors of a course can populate the pool of questions. Note that this

plugin is basically a supporting backend for other plugins, like the quizzes and exams

discussed above.

139

3.5 Tasks

Tasks are the objects handled by the tasks plugin. Apart from tautological, this state-

ment intends to mean that, in SocialWire, a task is a supervised activity, i.e., an activity

that unfolds during a given timeframe and receives feedback from the instructors. The

feedback has the purpose of heading the students learning in the right direction, paving

the way so that they can acquire the knowledge by themselves. Therefore, tasks differ

from questions in that the former are inherently sequential over time, and more impor-

tantly, in that tasks may have partial deliverables: pieces or parts of work that deserve

examination prior to the complete fulﬁllment of the task. Simply stated, tasks have a

history of development that ordinary answers lack.

For the same reasons exposed in other plugins, tasks can be used either to grade the

students or to give them some share of (ranking) points. Correspondingly, this plugin

does not work isolated, but it interacts with the ranking plugin and the gradebook. In

addition, tasks can be incorporated to the e-portfolio of the student, as a demonstration

of its abilities, its performance and knowledge of a topic. See Figure 4 for a screenshot

of the current implementation of this plugin.

3.6 e-portfolios

The concept of a portfolio aims at gathering all the outputs produced by students, what-

ever their form, contents or support, in a single place. When evaluated as a joint effort,

they provide qualitative information about the personal learning paths followed. Thus,

a portfolio offers a double value. On one hand, teachers are able to identify the whole

learning process that the student unfolds, with its achievements and pitfalls. On the

other side, the student can demonstrate to others his personal achievements in the dis-

cipline, showing the results of every activity conducted within the curriculum (projects,

examinations, complementary work, self-study, etc.).

In SocialWire, an e-portfolio is a container object that stores evidences about the

work a student did along a given timeframe. The evidences collected in his/her e-

portfolio are not limited, of course, to the grades received as a result of his/her learning.

It includes all types of deliverables produced as a result of participation in any of the

plugins described up to this point: questions and answers, tasks, challenges, even its

position in the ranking system. E-portfolios are individual, per student, since the notion

of an e-portfolio for groups does not make sense in this context. We emphasize that

e-portfolios are simple containers, the implication being that e-portfolios do not bear

direct relationship to a speciﬁc assessment procedure. In other words, how to valuate

a student’s e-portfolio is a matter pertaining to the examiner. To aid in this process, e-

portfolios can be linked or associated in some way to a rubric (see the next subsection),

which is a description of the rules to evaluate an evidence. It is not mandatory that this

association exists, however.

3.7 Rubrics

Generically, a rubric is a set of clear criteria to assess any learning activity, published

in advance so that the community of students is aware of the requirements their output

140

Fig.4. View of the tasks module.

should satisfy. Also, the rubric is a commitment to the instructors, who comply to judge

the outcomes of the activities according to the principles stated in that rubric. Clearly,

141

rubrics are particularly important in informal learning, in order to avoid any bias or

subjectivism when the quality of the students’ work is under evaluation.

In SocialWire, rubrics are implemented as a matrix of categorical dimensions, such

as clarity, correctness, extent, etc. For each dimension, the matrix speciﬁes a range and

a scale of achievement (not necessarily linear), along with an optional description for

every level in the scale (usually, the scale will be discrete, i.e., the range is divided

into suitable intervals or buckets). Otherwise, rubrics may be deﬁned freely, both in

evaluated dimensions and type of measurement scale. To glimpse an example, Figure 5

displays part of the screenshot showing a hypothetical rubric, as seen by the user.

Fig.5. View of the rubrics module.

3.8 Gradebook

This plugin is self-explanatory. The gradebook is a compilation with the grades a stu-

dent has received for his homework, exams or any other formalized test of knowledge.

In SocialWire, without loss of generality, gradesare always numeric, and the plugin pro-

vides functions to compute simple descriptive statistics for an individual or for a group:

averages, medians/modes, histograms, ... Obviously, the gradebook plugin receives in-

put from every other plugin where the students make their work, and does not generate

output to any other plugin. Thus, the design of the gradebook is largely independent of

the other system components.

3.9 User Ranking - Reputation

One of the main motivations for developing SocialWire was the belief that online social

interaction between students promotes the effectiveness of learning. In online systems,

where the users cannot hold face-to-face contact, the possibility to compare one’s per-

formance with that from other participants turns out to be of outmost importance for

sustaining the students’ engagement. One of the simplest and most effective ways to

push the dynamics in groups is to use an understandable rank of members. The rank

shows a number of things simultaneously. First, it exhibits the degree of achievement

for every users, i.e. where I am vs. where are the others. Admittedly, a list is an over-

simpliﬁcation for measuring degree of achievements, but the implicit information con-

tained in it is otherwise a good feedback that tells how well a student fulﬁlled the work

142

compared to his colleagues. Second, it establishes a hierarchy within the group. Once

known, the ranking operates unconsciously on the whole group: answers, outcomes or

deliverables contributed by highly-rated members will probably be better esteemed by

the rest. In the opposite direction, persons with high position in the ranking are likely to

be compelled to deliver high-quality work in future activities if they aim to keep their

rank.

The ranking plugin in SocialWire simply organizes all this information about the

points received by every group member. In this respect, recall that the points awarded

to a student in a particular task can come from the instructors or from votes from other

group members, depending on the type of activity. The plugin just accumulates the

ratings of each user and sorts the list appropriately. Whether the position in the ranking

will affect the ﬁnal grade of the student is an open issue, up to the instructors’ decision.

The plugin is designed to be agnostic about these concerns. Similarly to the gradebook

plugin, this one is fundamentally a collector module that receives input from any other

module in the system, organizes the information and renders the data in appropriate

form.

3.10 Polls and Decision Trees

The last plugin allows SocialWire to conduct polls. More generally, the plugin imple-

ments a voting system, where users can choose their preferred options among a set

of options, and about different questions (a poll may embrace several questions, re-

lated or not). Polls can be anonymous —SocialWire does not reveal the author of any

opinions— or not, and the system carries out the simple accounting of responses, mak-

ing a ﬁnal report with the data. A variant of a pure poll, also implementedin this module,

is that of a decision tree. It is well-known that decision trees are a common tool to solve

complex decision-taking problems, after following a series of individual decision steps.

Accordingly, in SocialWire a decision tree is essentially a sequential poll, where each

step in the poll is a single question. The logic for advancing is the natural one, namely

majority vote.

Polls and decisions are orthogonal to the rest of plugins. Although a decision tree

could be used to solve a problem (formal or informal) within a group, probably the most

realistic case will be to use it as a supporting tool in some of the plugins described so

far. Similar considerations apply to the polls, in the sense that those object have been

devised to take the opinions from the users, not to solve tasks by popularity.

4 Application

The tool has entered in a test stage during the previous academic year.

Initially it will be used in two subjects of the Communications Engineering Cur-

riculum:

– Computer Networks (second year, obligatory).

– Advanced Computer Networks (third year, in a speciality).

143

These subjects represent a coherent thematic unity and develop with different levels

of complexity the discipline of communications networks.

The sequential implementation will allow us to observe the response of our stu-

dents to informal learning activities and to detect the possible improvement (in working

habits, in participation) and the increase or decrease of previous social reputation.

Following, we summarize the methodology and results obtained the previous aca-

demic year in the subject Computer Networks.

These activities were carried out by means of the platform:

– Thirty questions proposed by students (opened during the whole term). Students

received 1 point for each approved question, 2 points for the best answer and 1

point for other good answers.

– Two questions proposed by teachers. We rated them with 5 points for the ﬁrst good

answer.

– A challenge. Students looked for a resource related to the subject (video, presenta-

tion, ...) for two weeks. After this time, each student distributed 5 votes among the

answers during a week. Finally, the three answers with more votes received 15, 10

and 5 points.

– A task. Programming a HTTP server (opened for a month).

– An assessed questionnaire (opened for a week).

– Practice tests during the two last weeks of the course.

The continuous assessment represented a 30% of the ﬁnal grading (the task sup-

posed a 10% and the questionnaire a 20%) and was not obligatory. The points obtained

by means of the questions and challenge modules supposed an extra 10% of the ﬁnal

grading (each student received a grading proportional to the points obtained).

From the results obtained, we highlight that 85% of the 54% of the students that

followed the continuous assessment and participated in the proposed games passed the

subject (and with the best grades). Nevertheless, only 30% of the students that did not

follow the continuous assessment passed the subject.

Other advantages of the informal learning activities proposed can be highlighted:

– Questions: The resolution of doubts and habitual problems helped to the joint com-

prehension of the subject.

– Challenge: Some students engaged in the search of interesting resources. Other

students had to read the contributions of the companions in order to vote and to do

reasoned comments.

This year we are repeating the experience, but we haveintroduced newactivities and

some changes in the methodology in order to motivate more students and to increase the

level of participation in these informal activities. We plan to report the results obtained

in the near future.

5 Conclusions

In this paper, we have described a software platform that sees online learning as a spe-

cial social activity process. In this realm, it seems natural to embed informally-deﬁned

144

activities into the design of the LMS, and moreover, to revolve the design around an

OSN instead of a LMS. That is precisely the novelty of SocialWire: start from an OSN

engine and adapt it to become not only an online social network, but also a learning

platform.

Preliminary tests of SocialWire in a real environment are being carried in our col-

lege, among undergraduate students. We are in the second year of the ﬁeld experiment,

and plan to report more results (degree of students engagement, participation, relation-

ship between the ranking/reputation and the distribution of grades, observations related

to the sequential implementation, etc.) in the near future.

As further work we are going to extend the functionality of the platform with new

modules, as:

– Forms module: This module can be useful for example for implementing activities

related to the detection of previous knowledge about a subject.

– Badges module: We think that badges assignment may generate positive emotions

and can help to motivate students.

Acknowledgements

This work has been partially funded by the University of Vigo, under its program for

innovation in education methodologies.

References

1. (http://coursera.org)

2. (http://edX.org)

3. Cross, J.: Informal Learning: Rediscovering the Natural Pathways that Inspire Innovation.

Pfeiffer (2006)

4. Johnson, S.: Everything Bad Is Good for You. Penguin (2005)

5. Lee, J., Hammer, J.: Gamiﬁcation in education: What, how, why bother? deﬁnitions and

uses. Exchange Organizational Behavior Teaching Journal 15-2 (2011) 1–5

6. Vassileva, J.: Toward social learning environments. IEEE Transactions on Learning Tech-

nologies 1 (2008) 192–214

7. (http://socialwire.es)

8. (http://elgg.org)

9. Festinger, L.: A theory of social comparison processes. Human Relations 7 (1954) 117–140

10. Sousa-Vieira, E., López-Ardao, C., Rodríguez-Pérez, M., Fernández-Veiga, M., López-

García, C.: Enabling social learning environments at the college level: A toolbox. In: IEEE

EDUCON’12, IEEE Global Engineering Education Conference. (2012)

145