Social Network System in Classroom: Antecedents of Edmodo © Adoption

Social networks provide various benefits to educational settings. Nevertheless, a dominant social network tool like Facebook is not suitable for classroom, due to lacks of privacy concerns. Edmodo is a private social network that is claimed to provide a secure learning platform for learners and educators. Although Edmodo is a Facebook-like tool, it has not yet prevalent for educational usage. Thus, this research’s objectives are to study antecedents of Edmodo adoption as a classroom collaboration tool, to compare the effect of antecedents, to explore university students’ views about Edmodo, using Thailand case. The results are applied to give guidelines for lecturers, to effectively apply Edmodo in their classroom. Studied factors consist of two perception factors, one instructor factor, and one student factor. Perceived usefulness, perceived ease of use, and instructor characteristics are the strong predictors of Edmodo acceptance.


Introduction
Web 2.0 has changed the way users consume contents.The main characteristic of Web 2.0 tools is users' active participation in the content of creation process.According to academic studies, the use of Web 2.0 in education is important to remain relevant and to meet needs of students in the twenty-first century (Ollis, 2011).Advantages of using Web 2.0 in education are creating new interaction styles between instructors and students, promoting students interaction outside classrooms, boosting collaboration on group projects, enhancing students' experience from active environments, responding students immediacy, sharing just-in-time contents to peers or students in other schools, and linking lecture information and assignments to various digital resources (e.g.blogs, RSS, multimedia clips, wikis, and internet resources) (Campos and Garaizar, 2010; ______________________________________________________________ _______________ Mathupayas Thongmak (2013), Journal of e-Learning and Higher Education, DOI: 10.5171/2013.657749.Mazman, 2009)).They are proven to enhance students' learning experience and to create many advantages in informal learning (Mirabolghasemi and Huspi, 2012;Potter, 2006).Online social networks are also confirmed to be effective teaching tools because most students already have accounts and their platforms are ready-touse (Towner et al., 2007).Comparing to traditional content management systems (CMS) tools, social network sites provide additional features that are media sharing, RSS, tagging, own brand and visual design, real time activity stream, groups, friends, and profile pages (Mirabolghasemi and Huspi, 2012;Brady et al., 2010).Some usages of social networks in higher education are library uses, faculty uses, and administrative uses for content generating, sharing, interacting, and socializing (Roblyer et al., 2010;Hamid et al., 2009).
In the case of e-learning in developing countries, Andersson and Grönlund (2009) do a critical literature review about comparing major challenges between developing countries and developed countries.The research identifies 30 challenges that are grouped into four main categories: course challenges, challenges pertinent to individuals' characteristics, technological challenges, and contextual challenges.Major challenges which are emphasized in developing countries are course, technology, and contextual challenges, whereas challenges related to individuals' characteristics have not yet been focused in developing countries.Bhuasiri et al. (2012) investigate the key success factors affecting e-learning systems in developing countries.The research specifies 6 dimensions of 20 critical success factors.Six dimensions are learners' characteristics, instructors' characteristics, institution and service quality, infrastructure and system quality, course and information quality, and extrinsic motivation.
For Thai context, Siritongthaworn and Krairit (2004) investigate the common construct of the students' use of e-learning.The result shows three main interactions that are human-to-human interactions, human-tonon-human interactions, and access duration.Siritongthaworn et al. (2006) identify factors influencing the level of elearning success in Thai universities.Three main drivers are characteristics of the organization, the instructor and the Internet environment.One key barrier is the student preference for instructor-led learning.Pagram and Pagram (2006) explore Thai perspectives about e-learning and propose the suggestion for Thai elearning designers.In sum, the result points that Thai educators should customize instructional design and development of e-learning materials to fit Thai students such as e-learning should be used as the supported tool rather than replacing classroom learning; chat, discussion groups, and video conferencing can provide the sense of community to online learners; and so on.Siritarungsri and Suwansumrit (2011) study the use of Webcasting to support graduate nursing students.Using this tool can lead to nursing learning achievements from giving students chances to build online social communities, share their health education knowledge with others, and gain writing and presentation skills.Social networks are not extensively adopted in the education filed as much as in other fields (Duncan and Chandler, 2011;Mazman and Koçak-Usluel, 2009).Some social networks tools were applied in teaching and learning such as Twitter, Twiducate, Facebook, Edmodo, and Ning (Mirabolghasemi and Huspi, 2012;Roblyer et al., 2010;Hineman and Norris, 2011;Forte et al., 2012;Galán, 2011)[15] (Brady et al., 2010) (Mack et al., 2007).Facebook is the most popular social network which reached one billion active users as of October, 2012.It also has many features including wall, pokes, news feeds, photos, etc.Even though, social networks, especially Facebook, has been widely criticized for privacy vulnerabilities (Roblyer et al., 2010;Campos and Garaizar, 2010;Galán, 2011).So, social networks dedicated to education such as Yammer, Edmodo, or Ning are suggested to be used (Campos and Garaizar, 2010;Galán, 2011).Edmodo is a private social platform which provides a secure space for teachers and students to connect and to collaborate (Duncan and Chandler, 2011;Halm et al 2012).It is easy to apply to classrooms since its appearance is similar to Facebook, that many students are already familiar with (Haefner and Hanor, 2012;Holland and Muilenburg, 2011).However, those students need to be made aware of what constitutes the social networks tool and to be suggested the opportunity to use it for meaningful purposes (Ng, 2012).
There are only few researches regarding academic usages and educational benefits of social networks in education (Mirabolghasemi and Huspi, 2012;Brady et al., 2010).For instance, relationships between friendship networks, advising networks, and adversarial networks and students' performance were studied by Yang and Tang (2003).Online discussion group was applied to be an additional tool in Cooper's class.Students' point of views, online group work is suitable for an accelerated course (Cooper, 2009).Mazman and Koçak-Usluel (2009)

Edmodo Adoption Constructs
The constructs in this study were adapted from a well-known model, TAM, and previous researches.The proposed model assumed that the dependent variable (Intention to Use) is affected by independent variables that are perceived usefulness, perceived ease of use, instructor characteristics, different types of student characteristics (Dependent/ Collaborative/ Independent), and past behaviour.Table 1 shows literature sources of each construct.

Table 1: Constructs in This Study and Their Literature Sources
Regarding acceptance constructs, Technology acceptance model (TAM), an extensively used theory in the information systems field, specifies two important factors influencing intention to adopt technology: perceived usefulness and perceived ease of use (Davis, 1989).TAM is later developed to TAM3 by Venkatesh and Bala (2008).TAM3 also specifies that both perceived usefulness and perceived ease of use are significantly related to new information technologies adoption in the workplace (Venkatesh and Bala, 2008).Perceived usefulness is the most important determinant of behavioural intention for all time period of information technology usages.It positively affects intention to use social network sites.Perceived ease of use has a positive influence on intention to accept general social network sites too (Sledgianowski and Kulviwat, 2008).Mazman and Koçak-Usluel (2010) indicate that perceived usefulness and perceived ease of use have positive effects on Facebook adoption in academic usage.The usefulness of education-based social network, Ning, is also confirmed by students in terms of allowing more frequent collaboration with peers and colleagues within a course, allowing them to communicate more effectively, more convenient than face-to-face classes for sharing and discussing ideas (Brady et al., 2010).Students also perceived the usefulness of using social networks for classwork in terms of convenient (Roblyer et al., 2010).Towner et al. (2007) confirm that students agree that Facebook is a useful tool for them and their class-related collaborations.
Therefore, the first hypothesis and the second hypothesis are proposed as follows: H1: Perceived usefulness is positively associated with the intention to use.H2: Perceived ease of use is positively associated with the intention to use.
Regarding instructor-related construct, Lahadi et al. ( 2012) emphasize roles of educators in establishing a clear purpose of social networks usage for students, encouraging students to respond to other students, updating materials and course topics, etc.Because students have a limited understanding of how technology could support their learning, instructors are

Constructs Sources
Perceived Usefulness (Venkatesh and Bala, 2008) Perceived Ease of Use (Venkatesh and Bala, 2008) Instructor Characteristics (Selim, 2007) Student Characteristics: Dependent (Charkins et al., 1985) Student Characteristics: Collaborative (Charkins et al., 1985) Student Characteristics: Independent (Charkins et al., 1985) Past Behaviour (Conner and Armitage, 1998) Intention to Use (Venkatesh and Bala, 2008) needed to know how to use the tools, systematically model their usage in classrooms, explicitly guide their students about the tools, and continue supports (Ng, 2012;Behnke et al., 2012).Volery and Lord (2000) summarize that instructor characteristics are one of key success factors in e-learning.Selim (2007) is identified that instructor characteristics in terms of instructor's attitude and control of the technology and instructor's teaching style are important for e-learning adoption.
Effective educator is the one who teaches the use of humor, stories, enthusiasm, and self-disclosure (Marzer et al., 2007) In addition, Instructor engagement is pointed to be one of key components in online courses too (Roblyer et al., 2010).Therefore, the third hypothesis is proposed as follows: H3: Instructor characteristics are positively associated with the intention to use.
Regarding student-related constructs, learning styles of students in terms of collaborative, independent, and dependent, affect learning and attitudes in the introductory economic course.Collaborative students are students who like classes with as many discussions as possible.Dependent students are students who like classes with lecture-based settings and prefer as many as guidelines from their instructors.Independent students are students who like classes giving opportunities to them to express opinions about courses' structures and contents (Charkins et al., 1985).In addition, different needs of learners are required different teaching styles to fulfil them.For instance, students who mainly want to get good grades, prefer instructors who help them to achieve their goals with low efforts.Students who have high intrinsic goals and low extrinsic goals, want instructors to put high demands on their learning, to encourage their critical thinking, and to ask for their self-studies and effort investments (Hativa and Birenbaum, 2000).Distance learning students favour independent learning styles.Dependent learners are more prefer on-campus classes than online distance classes (Diaz, 1999).Diaz (2000) also specifies that successful distance learning students (grade better than 'C') are the independent type.Therefore, the forth hypothesis to the sixth hypothesis are proposed as follows: H4: Student characteristic (Dependent) is negatively associated with the intention to use.

H5:
Student characteristic (Collaborative) is positively associated with the intention to use.

H6:
Student characteristic (Independent) is positively associated with the intention to use.
Regarding student-related constructs, past habit influences intention and behaviour in the theory of planned behaviour (TPB) (Conner and Armitage, 1998).Attitudebehaviour consistency is also affected by direct behavioural experience (Regan and Fazio, 1977;Regan and Fazio, 1978).Early et al. (1993) point the importance of past experience or past behaviour on shaping intentions.The level of knowledge or experience with negotiation support systems (NSS) can be beneficial for building the intention to adopt the system (Lim, 2002).Pre-existing experience with social networks can make students able to use Edmodo's discussion boards and move from a teacher-centred question to a student-to-student discussion smoothly (Holland and Muilenburg, 2011).Therefore, the seventh hypothesis is proposed as follows: H7: Past behaviour is positively associated with the intention to use.

Instrument Development
A questionnaire is developed based on the research model, by adapting constructs from literature sources as shown in Table 1.The questionnaire composes of 21 questions.First question to fourth question (USEF1-USEF4) measure perceived usefulness; for example, "Using Edmodo will enhance my learning efficiency", "Edmodo will be useful to me".Fifth question to eighth question (EASE1-EASE4) measure perceived ease of use; for instance, "I do not need so much time to learn how to use Edmodo", "Using Edmodo is easy for me".Ninth question to twelfth question (INCH1-INCH4) measure instructor characteristics; for example, "Teacher always encourages me to participate in the class", "Teacher pays attention to students such as giving suggestions, answering questions, etc...".Thirteenth question to fifteenth question (INTU1-INTU3) measure intention to use Edmodo; for instance, "If I can access Edmodo, I will use It.", "I will use Edmodo during these one or two weeks".All above opinions are asked with the question "What do think about the following statements?" and answer choices in five points Likert scale (1 = strongly disagree, 5 = strongly agree).Student characteristic (sixteenth question), preference of Edmodo features (seventeenth question), gender (eighteenth question), and frequent access device (twentieth question) are collected using nominal scales.Examples of student characteristic questions are "I prefer to mainly have lectures in the classroom.I like the teacher to set topics and to describe clear details of assignments to me" -Dependent, "I like learning with as many as classroom discussions and interactions.I prefer group projects or learning from case studies."-Collaborative, "I like to participate in determining the course content and structure.If any assignments are given, I prefer to set the topics."-Independent.
Student characteristics variables were later treated as dummy variables.Nineteenth question (PAST) measure past behaviour in ratio scale with the question "How long have you used other social networks such as Facebook?".Last question is an open question about suggestions from students about using Edmodo to enhance classroom collaboration.

Data Collection
Online surveys created in Google Docs, with a convenience sampling, were used to collect the data.Questionnaires were sent to students of _____, _____ University, who took the Management Information System (MIS) course which applied Edmodo as a tool for classroom collaboration.This course is an introductory course for undergraduate students of all majors, not only MIS major.Two hundred and twenty nine questionnaires were sent.A total of 182 questionnaires were collected (a response rate 79.5 percent).

Reliability and Factor Analysis
The survey instruments were tested to assess their construct reliability and validity.Cronbach's alpha of each construct was assessed.EASE3 and INCH4 were deleted.Internal consistency of all factors are high since all Cronbach's alphas are greater than 0.8.Cronbach's alpha of INTU1-INTU3 is 0.794.Factor analysis were applied to check convergence validity (factor loadings are greater than 0.5) and discriminant validity (items were loaded with the right factor) and to form constructs from survey items.Principal axis factoring method with varimax rotation was applied.Three factors with eigenvalues more than 1 were derived.All factors can explain 72.533 percent of the cumulative variance of ten items as shown in Table 2.

Test of Model
Pearson correlation analysis was performed among independent variables: perceived usefulness (USEFS), perceived ease of use (EASES), instructor characteristics (INCHS), student characteristics (Dependent -STCH1, Collaborative -STCH2, and Independent -STCH3), and past behaviour (PAST).If the correlation between predictors is between 0.80 and 0.90, such predictor should not be included for the multiple regression analysis due to multicollinearity (Gururajan and Gururajan, 2008).Since the correlation between predictors STCH1 and STCH2/ STCH3 is significant and relatively high (r = -.881,p < .01/r = .331,p < .01),STCH1 was excluded from further analysis.
Relationships between six predictors (PAST, STCH2, EASES, USEFS, INCHS, STCH3) and a dependent variable (INTUS) are then explored by multiple regressions (enter method).The multiple correlation coefficient "R" for six predictors as shown in Table 4 represents the combined correlation of these predictors with the dependent variable (R = .667).The adjusted R square (R2 = .426)indicates that 42.6 percent of the variations in the Intention to Use can be explained by combined adoption factors.

Tolerance and Variance Inflation Factor (VIF)
were assessed to check multicollinearity.Tolerance less than 0.2 or 0.1 and VIF greater than 10 reveal collinearity problems (O'brien, 2007).All independent variables pass multicollinearity analysis with tolerance more than 0.9 and VIF less than 1.07, as described in Table x.From the above result, three of six independent variables (perceived usefulness, perceived ease of use, instructor characteristics) were found  to be significantly contributing to the prediction of dependent variable (intention to use) with p-value less than 0.01.Most influential factors for Edmodo adoption as a tool for classroom collaboration are perceived usefulness (b = 0.593, p = 0.000), perceived ease of use (b = 0.241, p = 0.000), and instructor characteristics (b = 0.164, p = 0.005) consecutively.

Limitations, Findings, and Implications
Some limitations exist in this study.Since the sample in this study is limited to undergraduate students from single faculty and single university, the research needs to be later replicated to examine the generalizability of findings.Nevertheless, some interesting results which instructors can apply to promote Edmodo usage are summarized.This study strongly supports perceived usefulness and perceived ease of use of the TAM model.(Hineman and Norris, 2011).Moreover, he/she can build perceived usefulness by pointing out various benefits of Edmodo as described above, and can bring more perceived ease of use of Edmodo into view by training students who are unfamiliar with Edmodo.This is supported by Brady et al. (2010) that emphasize the need for training and support for the use of SNSs in educational settings for both instructors and students.More advantages of Edmodo that teachers can emphasize are using technologies in education can help students to prepare for their future jobs; students will have flexible work hours inside and outside the classroom; students will have chance to reach most updated information with Edmodo both from instructors and other peers; working in groups with Edmodo in a cooperative way will help students to share their experiences and ideas; students will be more social and communicative because of the group work; students will have chance to produce content and to manipulate the content which supports their self-efficacy (Elmas and Geban, 2012).Two variables were not statistical significance: student characteristics and past behaviour.
Collaborative and independent characteristics of learners are not supported since created activities in Edmodo were not customized for these particular groups.This also conforms to prior studies of Neuhauser (2002), Hunt et al. (2002), and Thongmak (2011).Past behaviour is rejected due to the variety and instability of those activities (Ouellette and Wood, 1998)

Conclusion
Web 2.0, especially social networks, can be more beneficial for other areas such as education than entertainment only.It can be used to support both distance teaching and to fulfil physical classroom learning.Applications of social networks in education generate a wide range of benefits such as new collaboration styles, enhancing modern classroom experiences, resource sharing in various formats, etc.So, this paper aims to study vital drivers for social networks' adoption.Edmodo is chosen because it is less known and less used even though it provides more secure and easy platform than a popular social network, Facebook.Technology acceptance model along with instructor factor and student factors are gathered to check their importance.Quantitative questionnaires were applied to reveal the results.The results show that instructors should emphasize the benefits of using Edmodo, educate students how to use some unfamiliar Edmodo's features, encourage online collaboration environments, and treat students with care.Edmodo's developers should also improve the tool's features to compete with other general purpose social networks.For further research, the acceptance of different educational social networks or within other environments should be study to generalize the results.Action research should also be applied to study online activities suiting for different groups of learners.

Table 5 : Result of Multiple Regression Model and Variance Inflation Factor Analysis Coefficients a
Other social network sites such as Facebook or Twitter should be applied together with Edmodo."[Respondent 57, Respondent 67] "Special interest groups should be set up for the benefit of people who are interested in those fields" [Respondent 15] "Edmodo should have features such as group creation, application development, etc." [Respondent 118] "I wish Edmodo has features to add friends because sometimes I want to privately chat with friends or instructor" [Respondent 58, Respondent 83] "Edmodo should notify ______________________________________________________________ _______________ Mathupayas Thongmak (2013), Journal of e-Learning and Higher Education, DOI: 10.5171/2013.657749.events more clearly, like Facebook does" [Respondent 75].Last, students expect Edmodo to have more users.They described that "Marketing strategies should be applied to encourage more