Factors which motivate the use of social networks by students/Factores que motivan el uso de las redes sociales por los estudiantes.
At the same time as these psychological and sociological studies, research has been carried out into the adoption of social networks in the field of education. Some authors (e.g., Donlan, 2014; Junco, 2015; Mazman & Usluel, 2010; Sanchez, Cortijo, & Javed, 2014) have analysed the motivation for the educational use of social networks. Both teachers and students recognize the benefits that social networks can bring as an educational resource. (Bicen & Uzunboylu, 2013; Hamid, Waycott, Kurnia, & Chang, 2015). Moreover, the introduction of social networks is associated with the most innovative methodologies which promote active and collaborative learning (Al-Kathiri, 2015; Long, 2015; Kim, Holman, & Gooreau, 2015; Wodzicki, Schawammlein, & Moskaliuk, 2012), improved classroom atmosphere, and group social cohesion (Asterhan & Rosenberg, 2015).
Manasijevic, Zivkovic, Arsic and Milosevic (2016) independently analysed the purposes of Facebook usage and its educational utility as a social network. The results confirmed the findings of previous studies which identified motivations of use associated with social development (Ellison et al., 2007), leisure activities and information (Sharma, Joshi, & Sharma, 2016). In terms of the educational purposes of Facebook usage, the results confirmed the ideas put forth by Mazman and Usluel (2010) and Sanchez, Cortijo and Javed (2014), which detected 3 dimensions: communicate, collaborate and share materials. Manasijevic et al. (2016) concluded their study by recommending research with structural models which can reveal the possible relationships between the purposes of use of social networks and educational motives. In this way, we could see how to integrate them effectively in educational contexts.
Research to date has dealt separately, in a fragmented manner, with factors associated with the use of social networks, and therefore, the available information about the possible interrelationships between them, is still limited. This makes it difficult to achieve the necessary holistic view that would help to better understand this phenomenon (Burrow-Sanchez, Call, Zheng, & Drew, 2011; Colas et al., 2013). In particular, from the conceptualization of social networks as tools for educational purposes, it is essential to identify their potentialities in order to improve teaching processes (Hamid, Waycott, Kurnia, & Chang, 2015), as well as to reveal those factors which make them attractive in order for them to be successful included (Mazman & Usluel, 2010). As a result, studies from an integrating perspective are required, addressing both motives for use of social networks and their analogies with educational purposes (Manasijevic et al., 2016).
In accordances with these approaches, a quantitative transversal study has been carried out following a model based on structural equations, the main aim of which is to analyse the motives that explain the use of social networks and especially their educational utility for students in the 4th year of Secondary Education.
Convenience sampling was performed. The final year of compulsory education was chosen to ensure that the age of the students was higher than the minimum allowed (art. 13 RD 1720/2007). Around 1,144 surveys from 29 institutions were collected out of a population of 1,792 students in the 4th year of Secondary Education attending 31 schools (both public and private) in the city of A Coruna. With a statistical level of confidence of 95% and in the event of maximum indeterminacy (p=q= 50 and K= 2) the margin of error was [+ or -] 1.74. In terms of gender, 47.2% (n= 540) were boys and 52.8% (n=604) girls. The students were aged between 15 and 18. The age distribution was 41.8% (n= 478) 15-year-olds, 45.2% (n= 517) 16-year-olds, 10.6% (n= 121) 17-year-olds, and finally, just 2.4% (n= 28) were 18 years old.
An ex post facto design based on the survey method was used (McMillan & Schumacher, 2010), applying a questionnaire of 251 items created ad hoc and organized in 5 thematic blocks related to Internet usage and social networks in adolescence. The initial instrument was validated by a panel of 5 experts in research methodologies and implementation of technology in education who evaluated aspects such as the uniqueness, relevance and importance of each item. A second version was also tested on a pilot group and, after making any suggested changes, the definitive questionnaire was designed.
In addition, the corresponding item-total correlation tests were performed, which indicate that the items have a homogeneous relationship with the scale they belong to. Cronbach's Alpha was used to measure reliability, which gave an overall result of .937.
For the current study, just 27 items dealing with social networks were taken from the questionnaire (see Table 1). The students completed a Likert-type scale using these five items; 1 (never), 2 (hardly ever), 3 (sometimes), 4 (nearly always), and 5 (always).
The questionnaires were given out by the researchers during school hours, having obtained prior consent from the school. They were previously informed of the purpose of the investigation and guaranteed confidentiality.
Anaysis of data
With the objective of responding to the subject under research, a multivariable analysis was carried out through the development of a structural equation model based on the Partial Least Squares technique (hereafter PLS). This technique is interesting not only for contrasting models from a solid theoretical basis, but also as a means for exploration (Barclay, Higgins, & Thompson, 1995), as in this study. Statistical treatment of data was done using SPSS 19 and Smart-PLS software.
As with any SEM methodology "Structural Equation Modeling", the use of PLS needs the development of a measurement model and a structural model (Tenenhaus, Vinzi, Chatelin, & Lauro, 2005).
A factorial analysis of the principal components of each construct was performed, in order to confirm that the indicators in each latent variable were one-dimensional.
To ensure relevance in the development of the factorial analysis, the variables were subjected to "Barlett s test of sphericity". The Kaiser-Meyer-Olkin index (KMO) was also used. Reliability was evaluated using Cronbach's Alpha test, as well as composite reliability. Authors such as Bagozzi and Yi (1989) say that the indices of composite reliability higher than .50 confirm the internal reliability of the construct.
Evaluation of convergent validity was performed via Average Varience Extracted (AVE). In order to analyse discriminant validity, the matrix of factor loadings and cross-loadings was obtained. Another criterion to verify the discriminant validity is the square root of the AVE for the construct being larger than the correlation between that construct and the others (Chin, 1998).
After having verified that the measures of the constructs were reliable and valid, the structural model was evaluated, analysing to what extent the predictor variables contributed to the explained variance of endogenous variables. In addition, [R.sup.2] was used to discover how much of the variance of endogenous variables was explained by the constructs which predict them.
In order to examine the stability of the parameter estimates offered by PLS, "Bootstrap" was used to calculate standard error in parameters, and Student t values. In order to evaluate the goodness-of-fit model, the proposal by Tenenhaus et al. (2005) was followed, through the application of the indicator "Goodness-of-fit" (GOF).
An identification of each construct making up the model is included in Table 1, together with reflective indicators taken to appropriately measure each of the latent variables specifying their mean score and standard deviation.
Results relating to the measurement model and its internal consistency, and key factors such as unidimensionality of constructs, reliability and convergent validity are shown in Table 2. Barlett's test of sphericity and the KMO index (Kaiser-Meyer-Olkin) supported the need for a factorial analysis.
In column A of Table 2, the Eigenvalue of the first two principal components are shown (after discarding indicators which did not promote unidimensionality). Column B of Table 2 shows the percentage of variance explained by the first two principal components. In this case, the first component is expected to be the one which explains most of the variance, which is in fact the case.
The high reliability of each of the constructs can be seen in column C and D of Table 2, being over .70, as suggested by Nunnally (1978) for early stages of research.
As column E in Table 2 shows, the AVE (Average Variance Extracted) in the five constructs is higher than the value of .50 recommended by Fornell and Lacker (1981). Another important issue to take into account is the aspect of the loads, through which the loads or correlations between the different indicators and their constructs can be seen. Following Falk and Miller (1992), the level of acceptance in loads has been set as greater than or equal to .505. As can be seen in column F of Table 2, this limit is exceeded.
Regarding the matrix of factor loadings and cross-loadings, the factor loadings were seen to be greater than the cross-loadings. That means that the indicators are more strongly correlated with their own construct than with others. Table 3 shows the correlation coefficients between the constructs, and the square root of AVE can be seen in the diagonals.
The modeling process was based on identifying the factors which explain motivation when using social networks, that is, the elements that drive teenagers to apply the different utilities of social networks. As a first step, a summary of the model proposed is shown in Figure 1.
After testing the reliability and validity of the constructs, the relationships between the variables which form the structural model were analysed. To do this, the [beta] coefficient (path coefficient) was calculated, which should be at least .20, according to Chin (1998). As can be seen in Figure 2, this assumption is confirmed by the relationship between the constructs "Versatility-Motives for use ([beta] = .621)" and "Versatility-Educational uses ([beta] = .394)".
With reference to the amount of variance in the endogenous variables which is explained by the constructs that predict them, the [R.sup.2] obtained was .496 for motivation of use and .152 for educational use. In both cases and following the criteria proposed by Falk and Miller (1992), the values are greater than .10.
To evaluate the degree of fit of the model, the index "Goodness of fit" (GOF) was used, the resultant value of which was .417, greater than the the value of .36 suggested by Chin (1998). Therefore, it can be said that the proposed model has good predictive quality.
The results are displayed graphically (Figure 2) in order to visualize the relationships between the different constructs of the model PLS more clearly.
The results show that versatility in the use of social networks is the most important variable and the one which has the greatest impact on motivation for using these technological tools by teenagers. Furthermore, the positive influence between versatility in the use of social networks and its educational usage is also meaningful. On the contrary, neither dangers nor worries arising from the use of social networks, or even their educational utility have a negative influence on the use of social networks.
Social networks are a very complex, changing phenomenon. For this reason, it is necessary to go beyond the numerical data shown by employment statistics, which -as suggested by Zheng and Cheok (2001), need to be regularly updated- to move towards a holistic view, which would let us assess the incidence of variables using social networks with educational purposes.
In response to this question, this study identifies 5 constructs, which groups 27 variables regarding the use of social networks. The construct "Motivation for the use of social networks" includes eight variables connected with the four components proposed by Notley (2009), and gathers the psychological, sociological and cognitive factors present in other studies, which as shown by Colas et al. (2013), analyse each of these aspects independently. One major difference to note is that this construct also contains variables related to specific characteristics of social networks (free-of-charge, speed) through which the individual perspective is increased and people gain added external factors which come into play in their position on technology.
The results show that versatility is the variable which is of greater importance when explaining the motives for using social networks. As Delgado (2013) points out, social network analysis allows us to connect micro-behaviours and macro-behaviours of the population to which it belongs. In this respect, "versatility", as an explanatory and predictive construct of motivation for the use of social networks, addresses issues which exceed the personal level and refers to a general perspective. The 7 variables included in the construct called "Versatility" point out the flexible character of social networks to facilitate, emphasize and enrich connections and relationships at different levels and for many purposes.
It also highlights the importance of interaction as a key element from which activities are enhanced, such as those related to information and communication, previously recognized in other research about social networks. Topaloglu, Cldibi and Oge (2016) conclude that the aims of social network users are to follow (people, news, events ...) and share (information, photos, videos ...). All of which are actions related to the presence of the individual in the social scene and their social influence (Cheung, Chiu, & Lee, 2011), actions which reinforce the links with content, individuals and/or groups, as far as their capital and well-being is ensured (Greenhow & Burton, 2011).
Another noteworthy result is that the variables included in the construct "Dangers and worries" associated with social networks do not have a negative impact on motivation for the use of social networks. It seems that young people are unaware of the risks arising from the misuse of social networks (Vanderhoven et al., 2014) or do not take precautions to face possible threats (Livingstone, 2008).
The five variables included in the construct called "Educational usage" have little influence on the motives for use. This reveals students' limited experience in the educational use of social networks. Therefore, their use is not explained for academic purposes. However, students recognize the potential of social networks because the construct "Assessments for school learning" is associated with motives for use and it should be viewed in a positive light. The binomial education-social networks is positively reinforced, if one considers that the results show that the versatility of the construct is, in fact, what has an impact on educational usage. So, the changing and adaptable aspect of social networks is one of the main attributes that enables their introduction and enhances their use in the educational sphere.
Duffi (2011) warns us not to bring social networks directly into the classroom and his recommendation to create a new scenography to place them didactically, is essential to deal with the transfer from the social sphere, where young people use social networks, to the formativecognitive sphere, which requires learning achievement. Besides, the integration of social networks in academic activities should be based on the characteristics of versatility highlighted in this study which include all the psychological, social and cognitive aspects to be taken into account in all learning processes (Ellison et al., 2007; Junco, 2015; Mazman & Usluel, 2010).
The analysed constructs and the relationships that have been found allow us to move towards a map of social networks from which their possibilities for education can be seen. As shown by other authors (e.g., Yang & Brown, 2013), the motives for use have a major influence on the type of activity done and the results.
Therefore, knowing the motives for use of social networks lets us understand the factors that could encourage young people to use these tools in their daily school activity (Sanchez, Cortijo, & Javed, 2014). It could also be good for the teachers by helping them engage students to use social networks in education, for example through collaborative learning processes (Lisette, 2014).
In conclusion, making "the features of the social networks help students improve their personal growth with active, creative and cooperative learning experiences and increased interaction with people" (Topaloglu et al., 2016, p. 355).
Al-Kathiri, F. (2015). Beyond the Classroom Walls: Edmodo in Saudi Secondary School EFL Instruction, Attitudes and Challenges. English Language Teaching, 5(1), 189-204.
Asterhan, C.S.C., & Rosenberg, H. (2015). The promise, reality and dilemmas of secondary school teacher-student interactions in Facebook: The teacher perspective. Computers & Education, 55, 134-148.
Bagozzi, P., & Yi, Y. (1989). On the Use of Structural Equation Models in Experimental Designs. Journal of Marketing Research, 26(3), 271-284.
Barclay, D., Higgins, C., & Thompson, R. (1995). The Partial Least Squares (PLS) approach to causal modelling: Personal computer adoption and use as an illustration. Technology Studies (Special Issue on Research Methodology), 2(2), 285-309.
Bicen, H., & Uzunboylu, H. (2013). The Use of Social Networking Sites in Education: A Case Study of Facebook. Journal of Universal Computer Science, 19(5), 658-671.
Burrow-Sanchez, J., Call, M., Zheng, R., & Drew, C. (2011). Are Youth at Risk for Internet Predators?: What Counselors Need to Know. Journal of Counseling and Development, 59(1), 3-10.
Colas, P, Gonzalez, T., & De Pablos, J. (2013). Young People and Social Networks: Motivations and Preferred Uses. Comunicar, 40, 15-23.
Cheung, C.M.K., Chiu, PY., & Lee, M.K.O. (2011). Online social networks: Why do students use Facebook? Computers in Human Behavior, 27(4), 1337-1343.
Chin, W. (1998). The Partial Least Squares Approach to Structural Equation Modeling. In G. A. Marcoulides (Ed.), Modern methods for Business Research (pp. 295-336). New York, NY: Psychology Press.
Christofides, E., Muise, A., & Desmarais, S. (2012). Risky Disclosures on Facebook. The Effect of Having a Bad Experience on Online Behavior. Journal of Adolescent Research, 27(6), 714-731.
Delgado, O. (2013). Correlating students performance with social networks use in teaching. Procedia-Social and Behavioral Sciences, 93, 16681672.
Donlan, L. (2014). Exploring the views of students on the use os Facebook in university teaching and learning. Journal of Futher and Higher Education, 35(4), 572-588.
Duffy, P (2011). Facebook or Faceblock: Cautionary tales exploring the rise of social networking within teartiary education. In J.W. Mark & C. McLoughlin (Dirs.), Web 2.0 based e-Learning (pp. 284-300). New York, NY: Information Science Reference.
Ellison, N.B., Steinfield, C., & Lampe, C. (2007). The benefits of Facebook "friends": Social capital and college students' use of online social network sites. Journal of Computer-Mediated Communication, 12(4), 1143-1168.
Falk, R., & Miller, N. (1992). A primer for soft modeling. Akron, Ohio: The University of Akron Press.
Fornell, C., & Lacker, D. (1981). Evaluating structural equation models with unobservable variables and measurement error: Algebra and statistic. Journal of Marketing Research, 28, 39-50.
Greenhow, C., & Burton, L. (2011). Help from my <<Friends>> Social Capital in the Social Networks Site of Low-Income Students. Journal Educational Computing Research, 45(2), 223-245.
Hamid, S., Waycott, J., Kurnia, S., & Chang, S. (2015). Understanding students' perceptions of the benefits of online social networking use for teaching and learning. The Internet and Higher Education, 26, 1-9.
Hayta, A. (2013). A study on the effects of social media on young consumers buying behaviors. European Journal of Research on Education, Special Issue: Human Resource Management, 65-74.
long, S. (2015). Extending social networking into the secondary education sector. British Journal of Educational Technology, 47(4), 721-733.
Junco, R. (2015). Student class standing, Facebook use and academic performance. Journal of Applied Developmental Psychology, 36, 1829.
Lisette, T. (2014). Social networking: A collaborative open educational resource. Computer Assisted Languaje Learning, 27(2), 149-162.
Livingstone, S. (2008). Taking risky opportunities in youthful content creation: Teenager's use of social networking sites for intimacy, privacy and self-expression. New Media & Society, 10(3), 393-411.
Kim J.H., Holman D.J., & Goodreau, S.M. (2015). Using Social Network Methods to Test for Assortment of Prosociality among Korean High School Students. PLoS ONE, 10(4), 1-14.
Manasijevic, D., Zivkovic, D., Arsic, S., & Milosevic, I. (2016). Exploring students' purposes of usage and educational usage of Facebook. Computers in Human Behavior, 60, 441-450.
Mazman, S.G., & Usluel, Y.K. (2010). Modeling educational use of Facebook. Computers & Education, 55(2), 444-453.
McMillan, J., & Schumacher, S. (2010). Research in Education: EvidenceBased Inquiry. Harlow: Pearson Addison Wesley.
Molero, M.M., Martos, A., Cardila, F., Barragan, A.B., Perez-Fuentes, M.C., Gazquez, J.J., & Gil, J. (2014). Uso de internet y redes sociales
en estudiantes universitarios [Use of Internet and social networks by university students]. European Journal of Child Development, Education and Psychopathology, 2(3), 81-96.
Notley, T. (2009). Young People, Online Networks, and Social Inclusion. Journal of Computer-Mediated Communication, 14, 1208-1227.
Nunnally, J. (1978). Psychometric Theory. New York: McGraw-Hill.
Sharma, S.K., Joshi, A., & Sharma, H. (2016). A multi-analytical approach to predict the Facebook usage in Higher Education. Computers in Human Behavior, 55, 340-353.
Sanchez, A., & Martin, A. A. (2012). Generacion 2.0 2011. Habitos de uso de las redes sociales en los adolescentes de Espana y America Latina [Generation 2.0 2011. Habits of use of social networks among adolescents in Spain and Latin America]. Madrid: Universidad Camilo Jose Cela.
Sanchez, R.A., Cortijo, V., & Javed, U. (2014). Students perceptions of Facebook for academic purposes. Computers & Education, 70(1), 138149.
Tenenhaus, M., Vinzi, V., Chatelin, Y.M., & Lauro, C. (2005). PLS path modelling. Computacional Statistics & Data Analysis, 48(1), 159205.
Topaloglu, M., Caldibi, E., & Oge, G. (2016). The scale for the individual and social impact of students' social network use: The validity and reliability studies. Computers in Human Behavior, 61, 350-356.
Vanderhoven, E., Schellens, T., & Valcke, M. (2014). Educating Teens about the Risks on Social Network Sites. An intervention study in Secondary Education. Comunicar, 43, 123-132.
Wodzicki, K., Schwammlein, E., & Moskaliuk, J. (2012). Actually, I Wanted to Learn: Study-related knowledge exchange on social networking sites. The Internet And Higher Education, 15(1), 9-14.
Yang, C.C., & Brown, B.B. (2013). Motives for Using Facebook, Petterns of Facebook Activities, and Late Adolescents' Social Ajustement to College. Journal Youth Adolescent, 42, 403-416.
Zheng, R., & Cheok, A. (2011). Singaporean Adolescents' Perceptions of On-line Social Communication: An Exploratory Factor Analysis. Journal Educational Computing Research, 45(2), 203-221.
Mercedes Gonzalez Sanmamed (1), Pablo Cesar Munoz Carril (2) and Isabel Dans Alvarez de Sotomayor (3)
(1) Universidad de A Coruna, 2 Universidad de Santiago de Compostela and 3 Universidad de Vigo
Received: April 19, 2016 * Accepted: January 26, 2017
Corresponding author: Mercedes Gonzalez Sanmamed
Facultad de Ciencias de la Educacion
Universidad de A Coruna 15071 A Coruna (Spain)
Caption: Figure 1. Graphic representation of the model. Hypotheses and relationships between latent and observable variables
Caption: Figure 2. Results of the structural model based on PLS
Table 1 Reflective indicators used for measuring latent variables Construct/ Reflective Description Latent variable indicators Versatility in the VU1 I use social networks to use of social networks communicate with former friends VU2 I use social networks to communicate with current friends VU3 I use social networks to follow actions or opinions of the people I am interested in VU4 I use social networks to send messages VU5 I use social networks to share music VU6 I use social networks to read commentaries and news VU7 I use social networks to see/ share photos Dangers/Worries D1 I am concerned that my parents scold me or get angry with me because of using social networks D2 I am afraid that social networks create dependency D3 I am worried about not having time to use all the social networks I am registered D4 I am worried about the risk of cyberbulling Educational usage U1 I communicate and/or share information and resources related to my classes with my colleagues in the centre U2 I communicate and/or share information and resources related to my classes with the students I know from other centres U3 I communicate and/or share information and resources related to my classes with students I don't know U4 I create groups to do the tasks, homework, classroom projects, etc. U5 By using social networks I help my classmates Valuations in school VS1 The use of social networks in learning each subject would make them more attractive VS2 By using social networks I learn different things VS3 Teachers should use social networks in their clases Motivations for using MU1 Social networks allow a fast social networks communication MU2 Social networks let me be permanently in contact with my friends MU3 I connect to social networks to be up to date MU4 I like to use social networks because the communication is free of charge MU5 Social networks are easy to use MU6 I connect to social networks because it is fun MU7 I registered on social networks to share photos and videos MU8 Through social networks I can make plans Construct/ Reflective Mean SD Latent variable indicators Versatility in the VU1 4.16 1.192 use of social networks VU2 3.59 1.352 VU3 4.02 1.228 VU4 4.18 1.087 VU5 4.29 1.118 VU6 3.76 1.206 VU7 4 1.169 Dangers/Worries D1 1.66 1.086 D2 1.99 1.246 D3 1.69 1.081 D4 2.29 1.419 Educational usage U1 3.1 1.358 U2 2.49 1.972 U3 1.69 1.147 U4 2.07 1.292 U5 2.68 1.247 Valuations in school VS1 3.34 1.374 learning VS2 3.3 1.265 VS3 2.76 1.338 Motivations for using MU1 4.18 1.087 social networks MU2 4.29 1.118 MU3 4.16 1.192 MU4 3.59 1.352 MU5 3.76 1.206 MU6 4 1.169 MU7 3.69 1.311 MU8 3.9 1.245 Table 2 Unidimensionality, reliability and convergent validity of indicators and model constructs Unidimensionality Constructs and (A) Eigenvalue For (B) Explained indicators the first and second variance for the component first and second component Versatility in the 4.334 .942 48.156% 10.468% use of social networks: VU1 VU2 VU3 VU4 VU5 VU6 VU7 Dangers/Worries: 2.027 .766 50.664% 19.143% D1 D2 D3 D4 Educational usage: 2.565 .827 51.296% 16.548% U1 U2 U3 U4 U5 Valuations in 1.996 .608 66.530% 20.272% school learning: VS1 VS2 VS3 Motives for using 4.971 .923 49.715% 9.230% social networks: MU1 MU2 MU3 MU4 MU5 MU6 MU7 MU8 Reliability Constructs and (C) Cronbach's (D) Composite indicators Alpha reliability Versatility in the .835 .876 use of social networks: VU1 VU2 VU3 VU4 VU5 VU6 VU7 Dangers/Worries: .674 .800 D1 D2 D3 D4 Educational usage: .760 .833 U1 U2 U3 U4 U5 Valuations in .747 .850 school learning: VS1 VS2 VS3 Motives for using .872 .899 social networks: MU1 MU2 MU3 MU4 MU5 MU6 MU7 MU8 Convergent validity Constructs and (E) AVE (Average (F) indicators variance extracted) (Loadings) Versatility in the .506 use of social networks: VU1 .639 VU2 .677 VU3 .618 VU4 .764 VU5 .698 VU6 .752 VU7 .813 Dangers/Worries: .501 D1 .722 D2 .738 D3 .738 D4 .629 Educational usage: .504 U1 .815 U2 .646 U3 .598 U4 .698 U5 .771 Valuations in .656 school learning: VS1 .857 VS2 .881 VS3 .678 Motives for using .528 social networks: MU1 .744 MU2 .745 MU3 .730 MU4 .707 MU5 .691 MU6 .765 MU7 .709 MU8 .724 Table 3 Coefficient of correlation between constructs Motives Dangers/ Educational for use Worries usage Motives for use (.711) * Dangers/Worries .188 (.707) * Educational usage .334 .268 (.709) * Valuations in .349 .161 .274 school learning Versatility .689 .235 .394 Valuations in Versatility school learning Motives for use Dangers/Worries Educational usage Valuations in (.809) * school learning Versatility .324 (.726) * * Square root AVE construct