Content uploaded by Dimitra Tsovaltzi
Author content
All content in this area was uploaded by Dimitra Tsovaltzi on Sep 19, 2016
Content may be subject to copyright.
Effects of Argumentative Knowledge Construction on Attitude Change
Processes in SNS-blended seminars
Paper
Abstract (150 words) (Arial, 10 pt)
New self-organized and large-scale forms of communication, like SNS (social networking sites), bring new
possibilities for supporting argumentative learning, that is learning through argumentation. Because of the
social character of the interactions in SNS, SNS may extend our knowledge on group processes and
outcomes of argumentative learning like attitude change. This article presents a comparison of different
supports for argumentative learning in a university teacher-trainee course on communication theory that
included weekly SNS discussions: Group Awareness Tools (GATs), to increase attitude conflict awareness,
vs. argumentation scripts, as a cognitive guidance to help learners capitalize on this awareness. We use
Social Network Analysis (SNA) to analyze conversational data and data from a communication attitude
questionnaire on group-level processes relevant to attitude change during argumentative SNS discussions:
number of interactions, information flow control, influence distribution, and attitude similarity. Both GATs
and argumentation scripts influence argumentative processes, but scripts influence more processes.
Extended summary (paper presentation: 1500 words)
a) Group Awareness Tools and Argumentation Scripts for Attitude Change through Argumentative Learning in
SNS
Communication competence can help teachers communicate well with students, parents, colleagues, and school
administrative staff. Teacher trainees have pronounced negative attitudes towards the need for communication skills
(Ihmeideh & Al-omari, 2010) and communication theory seminars are offered to change their attitudes and improve
their skills. However, attitudes tend to be stable (Erber, Hodges, & Wilson, 1995) and attitude change presupposes
long-term deep learning and conflict awareness (Erber, Hodges, & Wilson, 1995). Deep learning can be obtained
through argumentation which supports Argumentative Knowledge Construction (AKC). AKC is the deliberate
practice of elaborating learning material by constructing formally and semantically sound arguments with the goal
of gaining knowledge (Weinberger & Fischer, 2006), but also co-construction of opinions and attitudes (Andriessen,
2006; Asterhan, Schwarz, 2007; Baker, 2003; Felton & Kuhn, 2001, Sassenberg & Boos, 2003, Tsovaltzi, Puhl,
Judele, & Weinberger, 2014; Wenger, McDermott, & Snyder, 2002). The social character of SNS may leverage
attitude differences and lead to socio-cognitive conflict and attitude change beyond what is possible in purpose-
specific collaborative learning tools. On the other hand, public discussions encourage the focus on attitudes and may
rather reinforce private beliefs Lampert, Rittenhouse, & Crumbaugh (1996). To seize the social affordances for
attitude change, GATs can visualize covert information about the group processes (Buder & Bodemer, 2008), and
make attitude differences salient, which is a prerequisite of dissonance and attitude change (Festinger, 1957). In the
collaborative setting of SNS this may lead to a higher number of interactions in an attempt to collaboratively
understand and resolve differences, less centralized information flow control, distributed influence, and maybe more
similar attitudes. Scripts can help analyze lines of ongoing argumentation and model adequate argumentative
processes and have been shown to improve processes and outcomes of argumentative knowledge co-construction
(Weinberger, Stegmann & Fischer, 2010). They may, thus, increase interactions by helping discussants elaborate on
their differences, externalize their attitudes and put them under scrutiny (Larson & Nussbaum, 2008), and thus lead
to more distributed reciprocal influence and more similar attitudes. Therefore, attitude change may occur (Erber,
Hodges, & Wilson, 1995). Their combination may promote socio-cognitive conflict and information exchange that
can lead to a higher level of influence between discussants and attitude change.
Results from inferential statistics have indicated that socio-cognitive conflict and argumentation quality can
be supported through SNS discussions with incorporated GATs and argumentation scripts, and their combination
even more, to influence their attitude towards communication skills over longer-stretches of time (semester long)
(Authors, 2015). Can similar effects be observed through the interactions and the network of relations that attitude
differences represent when students argue in SNS?
b) Method
In two long-term 2×2 field-studies, with factors GAT and argumentation script, we used Facebook, a prominent
SNS, to complement face-to-face teacher training seminars on communication theory with online argumentative
discussions over 9 weeks. German teacher trainees (N=148) filled out a weekly case-based questionnaire with cases
from every-day social interaction scenarios in the school, to capture their communication attitudes. Scenarios
consisted of four Likert-scaled answers on how a teacher may assess the situation. The questions covered two
communication attitude dimensions: multi-perspective / flexible attitudes vs. goal-oriented / structured attitudes,
following Buder & Bodemer (2008) and Jermann & Dillenbourg (1999, 2002), rated on a scale from 0 to 6. Each of
six seminars was accompanied by one closed Facebook group, where students held argumentative discussions on
problem cases using communication theories, to foster a balance between multi-perspective / flexible vs. goal
oriented / structured attitudes.
Students in the GAT conditions saw a graphical visualization of the result of the attitude questionnaire
presented in a Facebook application. It depicted their communication attitudes of their group (Figure 1), to make
conflicts salient and increase socio-cognitive conflict (Jermann & Dillenbourg, 1999), leveraging the high
interactivity in Facebook and foster attitude change. Students in the argumentation script conditions had to “like”
the best argument. They then received weekly feedback that evaluated the epistemic (theoretical concepts and
relations) and the formal (reasoning and evidence) argumentation quality for the most “liked”, and for the
instructors’ favorite argument. Students in the control condition merely discussed in their Facebook group.
Figure 1. Group Awareness Tool in Facebook.
SNA Analysis and Results
To measure interactions and information flow, we recorded the number of outgoing posts by each member and the
number of incoming replies to every post. Attitude data comes from the communication questionnaire. We
operationalize our dependent variables in SNA based on measures of graph theory and network analysis (Freeman,
1978; Stephenson & Zelen, 1989; Wasserman & Faust, 1994; Ester, Kriegel, Sander & Xu, 1996), as depicted in
Table 1:
Table. 1
Dependent Variable
SNA Operationalisation
Number of interactions
The sum of incoming and outgoing interactions for each member of a group measured across
time (8 weeks)
Information flow control
(information centrality)
The proportion of total information flow that is controlled by each node (group member) in a
network. It uses the distance between two nodes, traversing the attitude values (Euclidean
distance over time) of all nodes that mediate these two nodes
Influence distribution of
individual attitudes on
the group’s attitude
(out degree centrality)
The collective amount of attitude change of a group across time that can be attributed to the
attitude of an individual member in a weighted network. The weights are assigned based on
the attitude change after each discussion using the Euclidean distance of the two dimensions of
communication attitude questionnaire.
Influence distribution of
the group on individual
attitudes
(in degree centrality)
The collective amount of attitude change of an individual across time that can be attributed to
the attitude of the group in a weighted network. The weights are assigned based on the
attitude change after each discussion using the Euclidean distance of the two dimensions of
communication attitude questionnaire.
Attitude similarity
(density based
clustering)
The magnitude of attitude similarity inside a group across the two dimensions of
communication attitude questionnaire over time: the Euclidean distance of the two dimensions
is used to measure density.
We hypothesize that, participant groups using the GAT, the argumentation script, and their combination
will engage in a higher number of interactions, as they will become aware of differences and externalise more
information about their attitudes (H1); they will exhibit less centralized control on the total information flow, due to
new communication paths in the group (H2); their individual influence on the magnitude of attitude change of the
group will be lower, due to the increased interaction and reflection thereof among members (H3); equivalently, the
group’s influence on the magnitude of attitude change of individuals will be lower, due to an even distribution of
people central to the network (H4); and finally, they will develop higher attitude similarity (H5), because the above
changes in group processes will lead to co-construction of attitudes.
d) Results
We found significant differences regarding the number of interactions (H1), F(5, 142) =192.28, p = .00. As
expected, the argumentation script, F(1,144) = 123.44, p =.00, ηp² = .46, and the combination, F(1,144) = 63.64, p =
.00, ηp² = .31, increased the number of interactions with strong effects. However, the GAT had no effect, F(1,144) =
1.43, p =.23, ηp² = .01.
There were significant differences regarding information flow control (H2), F(5, 35.35) = 172.86, p = .00.
Following our hypothesis, we found strong effects for the GAT, F(1, 144) = 21.69, p = .00, ηp² = .13, the
argumentation script, F(1,144) = 131.02, p =.00, ηp² = .48 and their combination, F(1,144) = 90.02, p = .00, ηp² =
.39. They lead to evenly distributed information flow.
There were significant differences regarding influence distribution of individual attitudes on the group’s
attitude (out degree centrality, H3), F(5, 60.42) = 14.38, p = .00. We found a weak significant effect for GAT, F(1,
144) = 5.56, p = .02, ηp² = .04, a strong effect for the argumentation script, F(1,144) = 33.57, p =.00, ηp² = .19, and
for the interaction, F(1,144) = 23.07, p = .00, ηp² = .14. They forged evenly distributed influence of individual
attitudes.
There were significant differences regarding the centralized influence of the group’s attitude on the
individual attitudes (H4), F(5, 78.03) = 144.47, p = .00. As expected, we found a significant strong effect for the
GAT, F(1, 144) = 20.26, p = .00, ηp² = .12, the argumentation script, F(1,144) = 122.39, p =.00, ηp² = .46, and their
interaction, F(1,144) = 84.09, p = .00, ηp² = .37. They lead to evenly distributed influence of the group’s attitude on
individual attitudes.
There was a significant difference regarding attitude similarity (H5), F(5, 21.9) = 9.33, p = .00. As expected, we
found a strong significant effect for the argumentation script, F(1,50) = 9.40, p = .004 ηp² = .16. On the contrary,
neither the GAT, F(1,50) = 0.11, p = .74 ηp² = .002 nor its combination with the script, F(1,50) = 1.88, p = .18 ηp² =
.036 fostered similarity.
e) Discussion
The results revealed that GAT and argumentation scripts, and their combination can cater for a more equal
distribution of information flow, of individual attitude influence on the group’s attitude and vice versa during
argumentative discussions in SNS. Interactivity was increased either by the script alone or combined with the GAT,
but not by the GAT alone; a possible indication that GAT takes over the role of externalization promoted by the
script. Moreover, the argumentation script helped learners change their attitude and attain similarity, which did not
occur with the GAT or their combination. This might indicate that members resolved conflicts together when they
received cognitive guidance by the script, or it might be a sign of increased group bias (Sassenberg & Boos, 2003)
without explicit awareness of attitude differences (Festinger, 1954).
The results of this study compliment previous results on attitude change and show that argumentative SNS
discussions as part of university courses can be leveraged through GATs and argumentation scripts to support
argumentative group processes that are relevant to attitude change. More research is necessary to reveal the
intricacies of how group bias is affected through GATs and argumentation scripts, and to investigate the usefulness
of such tools in the wild, which may have a large impact not only on university course learning, but on societal
development.
f) References
Andriessen, J. E. B. (2006). Arguing to learn. In K. Sawyer (Ed.), (Tran.), Handbook of the Learning Sciences (443-
459). Cambridge: Cambridge University Press.
Asterhan, C. S. C., Schwarz, B. B. (2007). The effects of monological and dialogical argumentation on concept
learning in evolutionary theory. Journal of Educational Psychology, 99(3), 626–639.
Baker, M. (2003). Computer-mediated argumentative interactions for the co-elaboration of scientific notions. In
Arguing to learn (pp. 47-78). Springer Netherlands.
Buder, J., & Bodemer, D. (2008). Supporting controversial CSCL discussions with augmented group awareness
tools. International Journal of Computer-Supported Collaborative Learning, 3(2), 123–139.
http://doi.org/10.1007/s11412-008-9037-5
Erber, M. W., Hodges, S. D., & Wilson, T. D. (1995). Attitude strength, attitude stability , and the effects of
analyzing reasons. In Petty, R. E. & J. A. Krosnich (Eds.), Attitude Strength: Antecedents and consequences
(pp. 433–454).
Ester, M., Kriegel, H. P., Sander, J., Xu, X. (1996). A density-based algorithm for discovering clusters in large
spatial databases with noise. Proceedings of the Second International Conference on Knowledge Discovery
and Data Mining (KDD-96). AAAI Press. 226–231.
Felton, M., & Kuhn, D. (2001). The Development of Argumentive Discourse Skill. Discourse Processes, 32(2),
135–153. http://doi.org/10.1207/S15326950DP3202&3_03
Festinger, L. (1954). A theory of social comparison processes. Human Relations, 7, 117–140.
Fischer, F., Kollar, I., Stegmann, K. & Wecker, C. (2013). Toward a script theory of guidance in computer-
supported collaborative learning. Educational psychologist, 48(1), 56-66.
Freeman, L. C. (1978). Centrality in social networks conceptual clarification. Social Networks, 1(3), 215-239.
Jermann, P., & Dillenbourg, P. (1999). An analysis of learner arguments in a collective learning environment.
Proceedings of the 1999 Conference on Computer Support for Collaborative Learning - CSCL ’99, 33–es.
http://doi.org/10.3115/1150240.1150273
Jermann, P., & Dillenbourg, P. (2002). Elaborating new arguments through a cscl script. In J. Andriessen, M. Baker,
& D. Suthers (Eds.), Arguing to Learn: Confronting Cognitions in Computer-Supported Collaborative
Learning environments (pp. 1–6). Kluwer Academic Publishers. Printed in the Netherlands.
Lampert, M. L., Rittenhouse, P., & Crumbaugh, C. (1996). Agreeing to disagree: Developing sociable mathematical
discourse. In D. R. Olson & N. Torrance (Eds.), Handbook of human development in education (pp. 731–
764). Cambridge, MA: Blackwell.
Larson, Nussbaum, E. M. (2008). Collaborative discourse, argumentation, and learning: Preface and literature
review. Contemporary Educational Psychology, 33(3), 345–359.
Puhl, T., Tsovaltzi, D., & Weinberger, A. (2015). Blending Facebook into seminars for practicing argumentation.
Computers in Human Behavior 53, 605-616. doi: 10.1016/j.chb.2015.04.006
Sassenberg, K., & Boos, M. (2003). Attitude Change in Computer-Mediated Communication: Effects of Anonymity
and Category Norms. Group Processes & Intergroup Relations, 6(4), 405–422.
http://doi.org/10.1177/13684302030064006
Stephenson, K., & Zelen, M. (1989). Rethinking centrality: Methods and examples. Social Networks, 11(1), 1-37.
Tsovaltzi, D., Puhl, T., Judele, R. & Weinberger, A. (2014). Group awareness support and argumentation scripts for
individual preparation of arguments in Facebook. Computers & Education, 76, 108-118,
doi.org/10.1016/j.compedu.2014.03.012
Wasserman, S., and Faust, K. (1994). Social Network Analysis: Methods and Applications. Cambridge, ENG and
New York: Cambridge University Press.
Weinberger, A. & Fischer, F. (2006). A framework to analyse argumentative knowledge construction in computer-
supported collaborative learning. Computers & Education, 46, 71-95.
Weinberger, A., Stegmann, K. & Fischer, F. (2010). Learning to argue online: Scripted groups surpass individuals
(unscripted groups do not). Computers in Human Behavior, 26(4), 506–515.
Wenger, E, McDermott, R & Snyder, W.M., Cultivating Communities of Practice, HBS press 2002
