Incident 1:
Convicted murderer, death row inmate, Premalal Jayasekara was given permission by the Appeals Court on the 7th of September to attend sittings in Parliament. He was convicted of murder by the Court on 30th July, a few days before the General Election. However, he secured a seat in Parliament obtaining over 142,000 preferential votes from the Ratnapura District. On the 8th of September, he gave oaths as a Member of Parliament.
Incident 2:
It was reported on the 8th of September that Prime Minister Mahinda Rajapaksa had proposed the ban of cattle slaughter in Sri Lanka and the decision was unanimously given the nod by the MPs representing the government. The decision has not been put to parliament yet.
The analysis below examines these two specific incidents to understand how they played out on social media, on Facebook in particular.
Limitations-
- All data collected belongs to public pages/groups on Facebook.
- The data extracted is through the results shown on the social listening tool Crowdtangle.
Methodology-
- Keyword-based searches were used to extract the data. A minimum number of keywords which are directly relevant to the issues were used to extract data specific to the above two incidents.
Overview of the two incidents:
Time Period One monitored: 12 AM on 8th September to 12 AM on 10th September which covers the 48 hour time period where both these incidents started.
The above graph shows that incident 1 has a less post count compared to incident 2. As such, compared to incident 1, incident 2 has 57% and 54% more posts and interactions respectively.
For the next portion of the analysis the author will use the top 50 most shared posts for each incident.
Incident One
The following is based on the contents of the posts.
Almost 3/4ths of the posts show a negative sentiment towards incident 1. It should also be noted that almost all the positive posts were generated through the official page of Premalal Jayasekara and his daughter, Senali Jayasekara’s page. Furthermore these two pages generated 9 out of the top 50 best performing posts.
These 50 posts were from 33 different pages. Out of these pages, page of MP Premalal Jayasekara ranks 31st and that of Ms. Senali Jayasekara ranks at 28th. Even though these pages rank at the bottom, posts belonging to these two pages amounts to 11% of the total interactions received for the top 50 posts.
However, it is important to analyse the interactions these posts received. The interactions for these posts may be high both because of a show of support or an expression of disagreement towards the content. To calculate this, the author has looked at the different types of reactions received by each post in order to get an understanding as to how the users perceived a particular post. This is solely based on the distribution of “reactions” received by each post and no comments have been analysed.
Incident Two
The following is based on the contents of the posts.
It can be seen that almost 2/3rds of the posts were in support of the ban on cattle slaughter.
Reach:
The maximum potential reach of the messaging related to incident 1 amounts to approximately 36.2 million while incident 2 amounts to 43.3 million.
Time Period One monitored: Here the author looks at the dataset from 12 AM on 10th September to 12 AM on 12th September. This will help understand whether any or both of the incidents continued to dominate the social media narrative in this 48-hour window and in what ways.
When compared with the first 48 hour period, incident 1 has resulted in post count loss of 66% while incident 2 has resulted in a loss of 56%. However it can be seen that incident 2 has 108% more post count compared to incident 1.
Reach:
During this time period, the maximum potential reach for incident 1 dropped to 14.6 million while incident 2 amounts to approximately 33.2 million, over two times the reach for Incident 1.
Observations:
It is observed that incident 1 died down after a peak which lasted a 48 hour period. This has been observed to be the case for most trending content on social media. However incident 2 dropped down marginally in the second 48 hour time period but managed to completely shadow incident 1 making it the most trending topic within the 96 hour time period. The interaction rates were high for both these issues although Incident 2 had a higher rate even with a high number of posts within the first 48 hour period. During the second 48 hour period the interaction rates were similar but the Incident 2 had the most posts by far.
Conclusion:
Over the course of our monitoring exercise, in most of the cases the trending topics on social media were standalone events, It is a rare occurrence that there are key overlapping incidents within such a short timeframe. The author is of the opinion that Incident 1 was shortly forgotten and shadowed with the introduction of Incident 2 as seen through the above data. It is to be observed whether this was the case of “suppressing” an important issue with another “new” issue by creating a “distraction”.
By: Prihesh Ratnayake, Social Media Analyst, Hashtag Generation
