Introduction
The information people choose (in the form of opinions, advice or ideas) determines to a great extent the knowledge they acquire (Barthelme, Ermine, & Rosenthal-Sabroux, 1998; Nonaka, 1994). However, in order to forego the costs of individual learning, people have evolved to acquire knowledge through social learning processes like teaching, language and imitation (Mesoudi, 2011). Specifically, some evolutionary scholars have focused on three biases that take place when grouped individuals interact: they tend to conform to the beliefs of the group (frequency-based); they tend to imitate the ideas of powerful or alike individuals (model-based); or they simply select information that is perceived as having more benefits compared to the other options (content-based), (Mesoudi, 2011; Richerson & Boyd, 2005).
These biases tend to happen whenever people are grouped, but never before have we seen as many individuals interacting as we do now. With almost half of the world’s population making use of the internet (Internet Live Stats, 2015) and given the amount of information that is being shared and received by users, online communities have had to implement different structures that simplify the sharing of information. However, at the same time that these structures simplify and tailor the information we need, they also make us prone to obtaining information that is biased (Kahneman, 2003; Tversky & Kahneman, 1981). These structures affect: the amount of information allowed to be transmitted to other users (maximum or minimum characters allowed per ‘post’), the type of information that can be used (for instance, text, images, or video), the reach within the whole online community (i.e., some opinions are shared only within a selected group of acquaintances while others are meant to be seen by any online user), and the level of conformity towards an idea shared by someone in the network (by making use of different rating-scales).
Objective
The main aim of this research will be to analyse how these different structures might bias the information people receive, and to determine which biases have a greater impact at the moment an individual is choosing from available opinions, advice or ideas. To achieve this, the current research done in social media was structured around the three group-biases (content, model, and frequency-based). The literature review showed that research has already being performed regarding what makes information attractive to others in terms of its content (Cheng & Ho, 2015; Cheung & Thadani, 2012; Jalilvand, Esfahani, & Samiei, 2011; Liu & Park, 2015; Park & Nicolau, 2015), and also in terms of online power or expertise (Iyengar, Van den Bulte, & Valente, 2011; Jacobsen, 2015; Litvin, Goldsmith, & Pan, 2008; Wu, Hofman, Mason, & Watts, 2011). However few studies have addressed the topic of conformity in online networks (Tsao, Hsieh, Shih, & Lin, 2015). Particularly, the differentiation between the personal and the total social network has been under-studied (Jiang, Ma, Shang, & Chau, 2014; Luo & Zhong, 2015). Moreover, although some research has also been performed regarding the comparison of rating-scales in online environments (Riedl, Blohm, Leimeister, & Krcmar, 2010, 2013), these studies have not differentiated personal networks within the online platforms.
Therefore, to address these gaps, the present study explores the following research questions: Which are the biases that mostly affect online choices? How strongly does conformity affect the choices in social media? Do people conform differently to the total-network than to their online personal-networks? Does the selection of information from someone’s personal network get affected by two different rating- systems?
Methods
To target the research questions, the study will adopt a quasi-experimental approach, and the data gathered will be both quantitative and qualitative, and longitudinal in nature. The quasi-experiment will consist of three years of data generated within an online educational website (PeerWise) where the participants will be (non-randomly allocated) undergraduate students of a particular module in the University of Sheffield. This module currently uses PeerWise throughout the semester, where students use it “to create [multiple choice questions] and to explain their understanding of course-related assessment questions and to answer and discuss questions created by their peers" (PeerWise, 2015).
The study will encompass three years of data: [1] The first year will have the characteristic that Peerwise users will be able to choose their usernames1 and rate2 each other’s questions from 0 to 5. [2] In the second year the change that will take place is that anonymity will be removed. This is, all students will be signed-in with their first and last names3, while the rating-scale continue to be 0 to 5. [3] Finally, during the third year students will continue to be logged-in with their first and last names, and the change will be that the rating scale will go from 0-5 to 0-1 (similar to a ‘like/dislike’).
Each year of the quasi-experiment will have 350 students (approx.) which will generate around 54,000 interactions4. This data will be analysed using statistical methods5. Moreover, at the end of each semester students will be asked to complete a questionnaire where their personal networks (within the group) will be mapped. The data from the questionnaire will then be compared with the way users interacted in PeerWise, using social network 6and sequence7 analyses. Finally, yearly focus groups will be used to get additional qualitative data that helps the researcher better understand the opinions and feelings of participants regarding the presence of their personal networks in online environments and the use of a particular rating scale.
Results
Theoretically, this research will add value by addressing the previously outlined research questions. Empirically, the research will create value by performing a real-life quasi-experiment which will enable to study conformity to personal-networks and comparison of online rating-scales with a novel methodology. Regarding practice and policy, it will help to better understand the application of social media to education, by studying which structures better enable students to obtain information and retain knowledge.
Future Work
This study is part of an ongoing Ph.D. At the time of the conference the researcher will be performing the first between-group comparison, and will therefore be able to comment on some of the preliminary results.
REFERENCES
Barthelme, F., Ermine, J.-L., & Rosenthal-Sabroux, C. (1998). An architecture for knowledge evolution in organisations. European Journal of Operational Research, 109, 414–427. http://doi.org/10.1016/S0377-2217(98)00067-8
Cheng, Y., & Ho, H. (2015). Social influence’s impact on reader perceptions of online reviews. Journal of Business Research, 68(4), 883–887. http://doi.org/10.1016/j.jbusres.2014.11.046
Cheung, C. M. K., & Thadani, D. R. (2012). The impact of electronic word-of-mouth communication: A literature analysis and integrative model. Decision Support Systems, 54(1), 461–470. http://doi.org/10.1016/j.dss.2012.06.008
Elzinga, C. H., & Studer, M. (2015). Spell Sequences, State Proximities, and Distance Metrics. Sociological Methods & Research, 44(1), 3–47. http://doi.org/10.1177/0049124114540707
Granovetter, M. S. (1973). The Strength of Weak Ties. American Journal of Sociology, 78(6), 1360. http://doi.org/10.1086/225469
Internet Live Stats. (2015). Internet Users. Retrieved May 18, 2015, from http://www.internetlivestats.com/internet-users/
Iyengar, R., Van den Bulte, C., & Valente, T. W. (2011). Opinion Leadership and Social Contagion in New Product Diffusion. Marketing Science, 30(2), 195–212. http://doi.org/10.1287/mksc.1100.0566
Jackson, M. O. (2008). Social and economic networks. Princeton, N.J.: Princeton University Press.
Jacobsen, G. D. (2015). Consumers, experts, and online product evaluations: Evidence from the brewing industry. Journal of Public Economics, 126, 114–123. http://doi.org/10.1016/j.jpubeco.2015.04.005
Jalilvand, M. R., Esfahani, S. S., & Samiei, N. (2011). Electronic word-of-mouth: Challenges and opportunities. Procedia Computer Science, 3, 42–46. http://doi.org/10.1016/j.procs.2010.12.008
Jiang, G., Ma, F., Shang, J., & Chau, P. Y. K. (2014). Evolution of knowledge sharing behavior insocial commerce: An agent-based computational approach. Information Sciences, 278, 250–266.http://doi.org/10.1016/j.ins.2014.03.051
Kahneman, D. (2003). A Perspective on Judgment and Choice: Mapping Bounded Rationality. American Psychologist, 58(9), 697–720. http://doi.org/10.1037/0003-066X.58.9.697
Litvin, S. W., Goldsmith, R. E., & Pan, B. (2008). Electronic word-of-mouth in hospitality and tourism management. Tourism Management, 29(3), 458–468. http://doi.org/10.1016/j.tourman.2007.05.011
Liu, Z., & Park, S. (2015
...
