הרשת היא המסר, תפקידם של קשרים חברתיים בהערכת אמינותם של תכנים מקוונים

Size: px

Start display at page:

Download "הרשת היא המסר, תפקידם של קשרים חברתיים בהערכת אמינותם של תכנים מקוונים"

Susanna Gordon
5 years ago
Views:

1 הרשת היא המסר, תפקידם של קשרים חברתיים בהערכת אמינותם של תכנים מקוונים מידע וטכנולוגיה 2017 ברשת האינטרנט ובארגון תכנים ויישומים ד"ר צחי חייט 10/5/2017

2 SOCIAL TIES AND CREDIBILITY ASSESSMENT 2

3 Credibility Assessment Credibility assessment is a key process in people s acceptance of information and in shaping their attitude towards information. 3

4 Three dimensions of credibility assessment Item characteristics User characteristics Interaction characteristics Credibility assessment 4

5 5

6 What are we actually looking at?

7 CREDIBILITY ASSESSMENT AND TIE STRENGTH ON FACEBOOK With Nicole Krämer and Stephan Winter, Universität Duisburg-Essen (Germany), and Arnon Hershkovitz, Tel Aviv University (Israel) 7

8 Credibility and tie Strength Tie strength is a combination of the amount of time, the emotional intensity, the intimacy, and the reciprocal services which characterize the tie (Granovetter, 1973). The strength of a tie is known to be related to information flow (Wirtz & Chew, 2002), and attitudes formation (Aral & Walker, 2012), hence it might have an effect on credibility assessment. 8

9 Source credibility Perceived Trustworthiness Perceived Expertise + = Perceived Credibility (Fogg & Tseng, 1999) 9

10 Procedure Experiment with undergraduate students 1) Facebook friendship 2) Stimulus 3) Outcome measures 10

11 Facebook Friendship In the first step, we invited 180 undergraduate students to become friends with a Facebook profile we created for this study. 11

12 Facebook Friendship Data was automatically collected, with the permission of our participants, using a designated script, coded for this study. The following data was collected about our participants friends: (a) the number of Likes and Comments they made on the participant profile, (b)how long have they been friends with the participant on Facebook, and (c) how many mutual friends they and the participant have. Based on this data, we calculate tie strength, inspired by Gilbert and Karahalios's model (2009). 12

13 Stimulus USE R IMA GE User Name Tuesday, February 2 at 2:34pm Deceptive Memories: Why The Human Memory is Susceptible to Mistakes Memories remain, but research has shown that they are frequently wrong: We tend to remember life in a more positive way, and sometimes even remember events that never happened. Brain researchers explain why this is not a major mistake of evolution, but rather a clever way of it to increase the human race sustainability. SCIENCE-NEWS.ORG 13

14 Source credibility Please answer the following question about [name of friend], your Facebook friend: [7 point Likert scale for agree/disagree with the following statements:] This person is well informed about [topic of the story] This person is honest 14

15 Evaluations of news credibility (Flanagin & Metzger, 2007) Based on five items: How complete is the information presented in this news story? How accurate do you find the information to be? How unbiased do you find the information to be? How trustworthy do you find the information to be? How believable do you find the information to be? Measured on a Likert scale 1-5 (ranging from not at all to a whole lot). Was used to create a content credibility index (Cronbach's alpha =.68) 15

16 Sample 124 undergraduate students completed the survey (69% respondent rate). The sample was primarily comprised of females (80.0%), with fewer males (19.7%) and other gender (0.3%) participants. Age of the sample ranged from18 to 29 years (M=22.20, SD=1.80). 16

17 linear hierarchical regression models: content credibility index Model 1 Model 2 b (SE) b (SE) Knowledge.03 (.06) -.02 (.06) Digital Literacy.06 (.10).07 (.09) Story Context.08 (.10).08 (.09) Tie Strength.31 (.10).44 *** (.05) Source credibility.37 ** (.04).43 ** (.07) Tie Strength X Source credibility -.25 *** (.23) N R Note: ***p<.001, **p<.01, *p<.05 17

perceived health outcomes What is the effect of the source? 3.5 3 2.5 2 1.

18 perceived health outcomes What is the effect of the source? Low Closeness High Closeness Low Knowledge in the field High Knowledge in the field 18

19 NETWORK DIVERSITY, AND TWEETS CREDIBILITY ASSESSMENT With Arnon Hershkovitz, Tel Aviv University (Israel) 19

20 And then there where more 20

21 Key question Dose the level of diversity within people who shared a given piece of information (e.g. family, friends, colleagues. etc.) on twitter, effect the recipient credibility assessment? 21

Complex contagion When information is risky, or controversial, the willingness to accept this information, and to further circulate it may require independent affirmation or reinforcement from

22 Complex contagion When information is risky, or controversial, the willingness to accept this information, and to further circulate it may require independent affirmation or reinforcement from multiple sources. This type of diffusion is referred to as complex contagions (e.g., Centola & Macy, 2007; Traag, 2016) because successful transmission depends upon interaction with multiple sources. 22

23 Procedure Online between subjects, experimental design with 188 undergraduate students 23

24 Participants twitter network (example) Only includes Twitter users that the participants follow and have mentioned or re-tweeted during the last year. 24

25 Division into clusters Clusters reveals the network underlying cliquish groupings of nodes that are highly connected between them, but poorly connected to other vertices. The community detection method used in this study is the spinglass algorithm 25

26 Mapping the Twitter networks Image taken from Prof. Bernie Hogan s work 26

27 Stimulus Is there a correlation between the diversity of the individuals who shared the content and the credibility assessment process? 27

28 Online survey 28

29 Variables IV - Diversity We tested our manipulation (by asking our respondents how diverse were the people sharing the content on a scale of 1-5). We found a significant difference in the scores for diversity in the experimental group (M=3.82, SD=1.13) and the scores of diversity in the control group (M=2.21, SD=.73) conditions; t(8)=2.89, p <.05 DV - Credibility: The credibility measures were found to be highly reliable (5 items; α =.73). 29

30 Findings Compering the rankings averages of both groups using the Wilxcon test Average ranking: Control group: 3.12 Experiment group: 3.7 Z=1.56; p<.01 Effect size, d= Reliabilty score High Diversity Low diversity 30

31 Conclusions These results highlight the role interpersonal ties plays in the process of credibility assessment. The network is the message? 31

32 32

How Behavior Spreads. The Science of Complex Contagions. Damon Centola

How Behavior Spreads. The Science of Complex Contagions. Damon Centola How Behavior Spreads The Science of Complex Contagions Damon Centola Financial Disclosures 2017 served as a consultant for CIGNA 2 Intuitions Which table is longer? 3 Intuitions It s a trick question 4