Research Article
Exploring Links between Online Activism and Real-World Events: A Case Study of the #FeesMustFall
Table 6
Significant changes for average negative sentiment.
| Date | Confidence interval (dd/mm/YY) | Confidence level (%) | From | To | Level |
| 07-08-2016 | (30-07-16, 08-08-16) | 99 | −0.42085 | −0.29499 | 7 | 06-09-2016 | (05-09-16, 06-09-16) | 100 | −0.29499 | −0.43423 | 6 | 08-09-2016 | (08-09-16, 08-09-16) | 98 | −0.43423 | −0.27649 | 9 | 19-09-2016 | (19-09-16, 20-09-16) | 100 | −0.27649 | −0.43862 | 5 | 10-10-2016 | (24-09-16, 13-10-16) | 97 | −0.43862 | −0.46454 | 7 | 01-11-2016 | (31-10-16, 14-11-16) | 99 | −0.46454 | −0.42244 | 8 | 22-03-2017 | (09-03-17, 23-03-17) | 95 | −0.42244 | −0.52847 | 7 | 29-03-2017 | (27-03-17, 05-04-17) | 98 | −0.52847 | −0.43271 | 8 |
|
|
(Confidence level for candidate changes = 50%, confidence level = 95%, CI = 95%, bootstrap = 10 000, without replacement, MSE estimates)
|