Academia.eduAcademia.edu

Figure 3 – uploaded by K. Hazel Kwon

See full PDFdownloadDownload figure
Figure 3. Original pass-along network compared with fragmented network when removing 10 positions with the highest between- ness centrality. Table 2. Actor-Based dynamic network analysis results (N = 317)

Figure 3 Original pass-along network compared with fragmented network when removing 10 positions with the highest between- ness centrality. Table 2. Actor-Based dynamic network analysis results (N = 317)

Related Figures (4)

Figure 1. The diagram of research design.
Figure 1. The diagram of research design.
Table 1. Descriptive summary
Table 1. Descriptive summary
Figure 2. Growth of celebrity campaign pass-along network: Each wave indicates a referral round. The big dot from (a) to (e) is Mark Zuckerberg; the big dots in the giant component in (f) are the highest actors in terms of reach centrality; the green actor (the big dot in the upper-right corner) is Demi Lovato with the highest out-degree centrality.
Figure 2. Growth of celebrity campaign pass-along network: Each wave indicates a referral round. The big dot from (a) to (e) is Mark Zuckerberg; the big dots in the giant component in (f) are the highest actors in terms of reach centrality; the green actor (the big dot in the upper-right corner) is Demi Lovato with the highest out-degree centrality.
Note. W = nomination wave; TC = t ratio for convergence; overall maximum convergence ratio was 0.254; **p <.01; ***p <.001 of these results, post hoc tests with standard residuals computed from observed and expected frequencies were conducted, revealing that the occupation category “other” was responsible for the disproportionate distri- bution between participants and nonparticipants. That is, celebrities in the remaining occupational categories were not significantly different in terms of the likeli- hood of participating in the Challenge. Race was found to be significant (6 = 12.81, t=5.14, p < .001), showing that White celebrities were more likely to participate in the Challenge than others. No difference was found in terms of gender. the individual sociodemographic effects, the signifi- cance levels of dyadic similarity effects were confident with large ¢ values. Specifically, celebrity recipients were more likely to accept the challenge from the celebrity senders who shared the same gender identity (6 =0.74, t=5.13, p<.001) or race identity (6 =0.58, t=3.09, p<.001) and fell into the same age group (6 =0.61, t=4.10, p<.001) as well as the same occu- pation category (S=1.88, t=12.68, p<.001) than from those who did not share such attributes. Occupational similarity showed a particularly large effect. These positively significant results suggest that it was homophily, and not heterophily, that influenced the growth of the celebrity nomination network dur- ing the Challenge campaign. of these results, post hoc tests with standard residuals
Note. W = nomination wave; TC = t ratio for convergence; overall maximum convergence ratio was 0.254; **p <.01; ***p <.001 of these results, post hoc tests with standard residuals computed from observed and expected frequencies were conducted, revealing that the occupation category “other” was responsible for the disproportionate distri- bution between participants and nonparticipants. That is, celebrities in the remaining occupational categories were not significantly different in terms of the likeli- hood of participating in the Challenge. Race was found to be significant (6 = 12.81, t=5.14, p < .001), showing that White celebrities were more likely to participate in the Challenge than others. No difference was found in terms of gender. the individual sociodemographic effects, the signifi- cance levels of dyadic similarity effects were confident with large ¢ values. Specifically, celebrity recipients were more likely to accept the challenge from the celebrity senders who shared the same gender identity (6 =0.74, t=5.13, p<.001) or race identity (6 =0.58, t=3.09, p<.001) and fell into the same age group (6 =0.61, t=4.10, p<.001) as well as the same occu- pation category (S=1.88, t=12.68, p<.001) than from those who did not share such attributes. Occupational similarity showed a particularly large effect. These positively significant results suggest that it was homophily, and not heterophily, that influenced the growth of the celebrity nomination network dur- ing the Challenge campaign. of these results, post hoc tests with standard residuals
Related topics:
Social Network Analysis (SNA)Interactive AdvertisingElectronic Word of MouthIce Bucket Challenge
580 California St., Suite 400
San Francisco, CA, 94104
© 2025 Academia. All rights reserved