I said I’d share the findings of my Masters dissertation so here goes. The post is a bit long but hey ho, I’m sure you’ll all cope.
The research investigates the use of social media by journalists and emergency responders in Strathclyde over three emergency events:
- the terrorist attack onGlasgowairport in 2007
- the Lanark school bus crash in early 2010
- the severe weather in December 2010.
The Co-op building fire and the severe storm in late 2011 were also tracked live.
The most interesting parts for me were the case study interviews and comparing the influence of emergency responders and media outlets.
I’ve uploaded the complete dissertation but this is a summary of the conclusions of the latter chapters.
Case study interviews
As can be seen by the comments from both Strathclyde Police and South Lanarkshire Council there has been a slow realisation that social media can be used by emergency responders to communicate directly with the public instead of the traditional way of using the media as a conduit for information.
It can also been seen that the media is also beginning to respond to the public’s use of social media and a recognition that sites such as Facebook and Bebo are ways of sourcing photographs and Twitter can be a source of concise comments to add to news stories, both online and in print.
However, for both the media and emergency responders, from 2007 to 2010 there seems to have been little appetite to learn how to use social media effectively – learning seems to have been driven by events, with little or no sharing of best practice, either within each group or across groups.
This is consistent with the findings of the Riots, Communities and Victims Panel (2011) which advised a review of how councils and the police use social media and an assessment carried out as to how they could improve.
It is also similar to the research undertaken by Bruns et al (2012) around the Queenslandfloods of 2011 where they showed that both groups had a role to play in emergencies: emergency services are the trusted source of information but the media is best placed to collate emergency information.
The case studies, together with the questionnaire results also show that there is a propensity to favour Twitter over other social media channels in gathering, broadcasting and sharing emergency information which is again borne out by Bruns et al (2012, p.12), mainly due to the ‘retweet’ function but also due to the difficulty of managing conversations on Facebook which can ‘rapidly swamp important information’.
The live tracking of these two events shows a maturation in the use of social media by Strathclyde Police, South Lanarkshire Council and STV Glasgow. By using mainly Twitter as a daily tool for communication they are building relationships, not only with their followers, but also with each other.
However, as shown by the Evening Times, many media organisations may still have a long way to go, both in learning how to use social media effectively as a day-to-day communications tool but also as a news gathering tool. It remains to be seen if news organisations who are already comfortable using social media will share their experience with newcomers or if they will be left to learn ‘on the job’ as an incident breaks, as their predecessors did.
Hashtags too present a further difficulty, especially for emergency responders. Often a hashtag has been created by the public before the emergency services are on the scene and decisions about whether to use the public hashtag or create a new one could affect how quickly and effectively essential information is disseminated. In their research Bruns et al (2012, p.29) go so far as to suggest that ‘messages should be designed to be passed along easily . . . and should contain hashtags relevant to the topic’.
The online tool Followerwonk shows that even although STV Glasgow and The Evening Times have relatively few followers, they still have significant influence.
This would suggest that a little more discernment around who organisations follow back or who they chose to follow could increase their influence.
Overall these diagrams show that the audiences for South Lanarkshire Council, Strathclyde Police and the area’s media organisations tend to be different, as do the audiences of the different media outlets themselves.
This would suggest that were they to work together on unified messages during an emergency instead of working autonomously, the information would reach many more people and the likelihood of messages ‘going viral’ would increase.
This work should be done during ‘quiet’ periods of learning and not during an incident when the message itself is the most important consideration, rather than refinement of the medium.
This is in keeping with the findings of the Riots Communities and Victims Panel (2011) which found councils and the emergency services use of social media weak and Bruns et al (2012, p.31) who consider the retweeting of official information as ‘amplification’ of key messages.
It would appear that news organisations are at different stages of using social media but that the level of its use is not relative to circulation figures, but more on how interested, individual journalists are in using it. Both the questionnaire and the case studies highlighted that it is often one journalist who uses social media, often starting with a personal account, but that using social media during a crisis such as a weather event or a major incident proves its worth to senior management and it is then developed in a more formal way.
Social media strategies and policies seem to have grown organically with the early adopters within Strathclyde’s media. However the laggards may be putting rules and regulations in at an earlier stage of its use compared to the early adopters.
The members of the Strathclyde Emergency Co-ordination Group (SECG) are monitoring and using social media in their normal day-to-day business as councils, NHS Boards, police and fire and rescue service. However during an emergency incident the approval process for any information coming from the Communications Group means that its release can be slower than a social media audience would expect. This is similar to the findings of Kapucu (2006) who found that hierarchical structures struggle in emergency situations when quicker problem solving and information sharing are required. It would also appear that there are no protocols for sharing information between members across social media platforms. Information sharing by retweeting amplifies important information and is critical during emergencies (Bruns et al, 2012).
During an emergency, Strathclyde’s news organisations will use information, video and photographs provided by the public on social media channels, if they can trace it back to source. This would suggest that the public is aware when a news organisation has used their content but it remains to be seen if this is still the case if there is ever a large-scale incident in Strathclyde which creates a high volume of social media content across different platforms. All interviewees described busy news rooms with tight deadlines where social media is an add-on for existing staff.
However, the usual journalism practice would appear come into play at the scene of an incident where there have been casualties or deaths, in that reporters will check the names of victims, casualty figures and other details with the lead emergency responder, rather than depend on information from the public. News organisations are however, comfortable taking photographs from a personal social media account once the person’s name has been officially released. As stated previously the ethics surrounding digital media are still being researched and with a constantly changing landscape ethics policies will need to keep up (Ess, 2009).
Strathclyde Police admitted that they are in the early stages of monitoring social media at an operational level and it remains to be seen how this would connect and connect with the other information that feeds into the SECG Communications Group cell during an emergency. This has been highlighted as critical to the flow of communication and the potential to correct misinformation by the Riots Communities and Victims Panel in their interim report (2011) following the London Riots.
Some local authorities are using social media to have two-way conversations and monitor situations and for them social media could be an effective communication tool during an emergency. However, other local authorities are still using social media as a broadcast tool which goes against the advice held in the SECG Communications Plan where agencies should identify their key audiences and ensure ‘effective two-way communication’ (SECG:PCG, 2011 p.32).
A turning point for emergency responders was the severe weather in 2010. Up until that point they had felt their duty was to inform the media during an emergency and that it was the media’s duty to inform the public. However, during the severe snow, with people trapped in cars, emergency responders realised that they could warn and inform the public themselves, although there was a realisation that the tone and content of the messages wasn’t quite right. By the time of the storms in 2011/12 this had been improved upon and emergency responders were also beginning to retweet each other’s information, although again there is no official protocol in place to make this a slicker operation.
Overall, during the time scale studied, both journalists and emergency responders have learned how to monitor and use social media during emergency events, rather than preparing during normal business. This goes against the advice of the Deputy Chief Constable of Tayside Police who advises that ‘you shouldn’t learn about social media while the bricks are flying and the cars are burning’ (Scobbie, 2012). There has been very little sharing of good practice, either within each group or across groups.
Twitter seems to be the platform of choice for both groups when it comes to putting out or sharing information, although journalists will use other social media platforms to glean pictures from personal accounts in certain circumstances. This is consistent with the findings of Bruns et al (2012, p. 31) who describe Twitter as:
deeply embedded in the broader media ecology, both drawing on and rapidly becoming a primary source of information for more mainstream news and media outlets.
By watching the storm event and the Co-op fire live this study shows that some organisations’ social media use is maturing and, not only is it being used as a communication tool for day-to-day business, it is beginning to be used to build relationships with their own audiences and each other. However, others still have a long way to go on their social media journey but not only are they playing catch-up, they should be aware of the speed at which the social media landscape can change and also that we have a generation of digital natives who take social media for granted and who expect organisations to communicate with them using it (Shirky, 2008).
Hashtags on Twitter seem to be a cause of concern for emergency responders and #hurricanebawbag showed them that, not only do they not control hashtags, when they try to control them it can have a detrimental effect on the flow of important information. Hashtags are essential for effective retweeting and as has been said earlier rewteeting amplifies any official messages (Bruns et al, 2012).
This research has also shown that each organisations’ Twitter account has very different sets of followers – they have very few common followers. Were they to share information more effectively, official information would travel further and quicker than it does at the moment. Also a little more discernment in who they follow could result in higher influence scores and a better flow of information if those influential followers choose to retweet official information.
This research has focused on the Strathclyde Emergency Co-ordination Group and it would be useful for research to be done across the other emergency areas to see if the results are consistent. However, the results here are consistent with research done inLondonfollowing the English riots in the summer of 2011 (Riots Communities and Victims Panel, 2011) and research done inQueenslandfollowing the severe floods in 2011(Bruns et al, 2012). This would suggest that the following high-level conclusions can be drawn for emergency communications in general:
- communications groups working during emergencies need to be aware of the expectations of a public used to the speed of social media. As discussed by Kapucu (2006) extreme events require quicker problem solving and information sharing than at other times
- there also needs to be a connection between operational monitoring of social media and the communications group’s use of social media to share information
- Twitter should be the main social media channel for communication information and conversation during an emergency event
- the effective use of hashtags and retweeting between emergency responders, the media and the public are essential for the amplification of key emergecny information.
This research indicates that social media has an important role to play for both emergency responders and the media but most important group of all is the public who need clear and consistent information during a crisis.
The following suggestions should be taken into consideration to enable emergency responders and the media to use social media effectively not only in their day-to-day business but also together during an incident.
Each Category 1 organisations should set up a Twitter account, as should media organisations. This should not be done during an incident but rather during ‘quiet’ periods when users have the time to learn how to use it effectively. Fail early so if mistakes are made, they are done while the account has few followers.
Twitter lets users create lists of important accounts which means the ‘noise’ can be filtered out from the rest. Lists of the following could prove useful during an incident:
- Category 1 responders such as blue light services, local authorities and health boards
- Category 2 responders such as airport operators, utility companies and harbour authorities
- Local/regional media
Community of Practice
The SECG model has been shown to work although there should be cognisance at a strategic level that the sign-off process for information can be too slow for the expectations of a public now used to the speed of social media. However, there is a disconnect between the SECG and the media in the area. Having the media attend SECG meetings during an incident is unrealistic but an online community could be set up. Public sector organisations are already comfortable using the Local Government Association’s online networking platform, the Knowledge Hub. An emergency response community should be created where best practice can be shared amongst, not only emergency responders but also the media.
Hashtags should also be discussed on the community of practice as early as possible during an incident. The hurricanbawbag hashtag was trending inScotlandby mid-morning on the day of the storm but had there been a platform for the media and emergency responders to discuss what hashtag was going to be used for official information there would have been less confusion and the media could have shared the official information in a more effective way.
Both the questionnaire and the case studies showed that media organisations and emergency responders are at various levels of social media use maturity. A buddy system should be set up where a more seasoned user is paired up with a relative newcomer. This would add an extra, one-to-one dimension to the community of practice and would help get everyone onto a similar level of use more quickly than if it were to grow organically.
During a crisis there is little time for either the media or the emergency communications team to sit on Twitter looking for information to retweet. However, there are tools available to automate retweeting and reposting. Tools such as TwitterFeed and If This Then That allow users to set up a rule that will recognise a hashtag being published on a named account which will then automatically retweet on the user’s account. Therefore if an official hashtag is agreed by the SECG the media can set up a rule that will retweet everything with that hashtag from Strathclyde Police, therefore the people who follow the Daily Record will see the official information coming from Strathclyde Police. These rules can be set up in advance during ‘quiet’ periods so that only the hashtag needs added. Retweeting not only amplifies the message (Bruns et al, 2012) but, looking at the Followerwonk data, the message will also travel to a wider audience by sharing it across accounts.
Applications such as Ushahidi and Sahana should be trialled to gauge their usefulness in the Strathclyde context. Ushahidi is an open-source crowd mapping tool and it was used to map fuel shortages in the run up to a threatened tanker drivers’ strike in March 2012. The Fuel Shortages Map allowed categories such as fuel shortages, deliveries and queues to be placed on a map of the country, either by tweeting using #fuelwatch or by filling in an online form. Volunteers around the world plot these pieces of information on the map (Fuel Shortages Map, 2012). The application has also been used for large scale events such as the severe snow inNew York in 2008, the Japanese tsunami in 2011 and the on-going political unrest inSyria (Ushahidi, 2012). A demonstration by a UK volunteer would show not only the power of the tool itself but also the power of the crowd. Again this should be done in a ‘quiet period’ rather than learning how to use it during the pressures of a crisis.
Tweet ups are informal events where people who use Twitter meet up face-to-face, often for the first time. Talking to other people who use social media tools but who are from sectors other than the media or emergency responders can be beneficial in that they may be using social media in different ways that could be applicable during an incident. Emergency responders and the media should be encouraged to attend these and even to organise their own.
So there you have it. A year of my life condensed into a blog post. I hope it was worth it!
And here’s the full thing – From John Smeaton to #hurricanebawbag