(Update: This post has been corrected due to my mistaken interpretation of Google Scholar search results – thanks to Deborah Lupton for the alert)
A Google Scholar search for academic articles published since the start of 2012 mentioning Twitter yields many pages of results.
The same search, but for articles relating to Twitter and health, also gives many results, including:
• a study investigating how Australian health organisations are using Twitter
• an article on mining Twitter for public health data
• and a study looking at use of Twitter to promote health literacy.
Meanwhile, “the power of Twitter” is the title of the email below that public health advocate Professor Simon Chapman recently sent to colleagues at the University of Sydney:
“When the Newtown gun massacre happened on Dec 14, I tweeted a link to a paper that Philip Alpers, Kingsley Agho, Mike Jones and I published in 2006, looking at what happened to gun deaths in Australia 10 years after law reform post Port Arthur.
In the month of December 2012, the paper got an amazing 82,310 hits & downloads. In the previous 6 years, it had received 16,974 hits.
I have 2,112 twitter followers, few of whom are gun control people. Twitter is an amazing vehicle to get your work in front of potentially thousands of interested people.”
Meanwhile, in another example of the power of Twitter for public health advocacy, this @GunDeaths account aims to tweet “every gun death in North America regardless of cause and without comment. Help us tell the story behind the statistics”.
It is a partnership with the online magazine Slate to provide “an interactive, crowdsourced tally of the toll firearms have taken since Dec. 14.” The data crowdsourced by the campaign is freely available for others to use.
It’s not hard to think of other ways this sort of approach could be used for public health campaigns.
And on other health-related uses for Twitter, below is a dynamic representation of a Twitter patient community, providing analysis of the conversations within the #rheum community last August (discussing all matters rheumatology-related).
[youtube]http://www.youtube.com/watch?v=CHmqsqnUp8Q[/youtube]
Each second of footage represents about six hours. According to the video blurb:
“A pink node represents a member of this community as they are participating in the conversations or are being pulled into the conversations by being referenced or mentioned. The larger the node, the more central or influential that member is to this community.
A green link between two pink nodes signifies a conversation, a direct interaction between the two as they happen. The thicker the link, the stronger the conversation or interaction is between them.”
A related article by a “simplifier” at the healthcare social media consultancy Symplur, Audun Utengen, aka @audvin, says:
“Twitter is in one sense an unstructured madness. But just like in the chaos of real life, people create meaning and structure within the noise. Rules are few and are made up as you go. Communities are created in a spur of the moment. Some last, others fade away.”
Is there a more dynamic space in public health and social debate?
Having been an extremely early adopter of twitter I now rarely use it in any shape. I remain militantly lukewarm about it at best.
As far as Chapman’s figures go.. .my long experience with web/net metrics is that one has to be extremely careful about just using one or two measurements.Especially hits, pageviews, etc. In addition its not uncommon these days for an individual to access the same thing from 4 different addresses, work – several, home, tablet, wifi in cafes, phone etc.
Downloads is a good measurement.
But I don’t find it particularly surprising that Chapman’s page had 4 times as much access since the Newtown massacre as in the last 6 years. I’m curious – was it actual download figures or page views or hits?
For some reason, perhaps a slow build up over years, there has been much mention and interest in the Australian Gun Control experience this time around in USA.
If you bang Australian Gun Control or similar into Google the first link up (other than news) is the Wiki entry. The wiki entry mentions Chapman et al in the body and has a link to the paper in the footnotes/references.
It seems to me its entirely plausible that the extra volume of downloads could have happened without twitter.
Despite all the excitement about twitter my experience over 5 years or is that the amount of noise makes it impractical unless you follow only 1 or 2 or 3 tags/streams.
NB: I’m generally in favour of Gun Control especially here in Australia and applaud Chapman’s work. (I might add I have owned guns and won shooting competitions) I can say that having been in the states, I have absolutely no idea what they can do about it there or about school shootings which seem unique to USA.
I think Melissa raises some good points about the potential of reusing online data for disease and public health policy surveillance.
However, it’s also true that this terrain has been well ploughed in recent years by authors such as Gunther Eysenbach.
Eysenbach has even coined a couple of neologisms to describe this field – infoveillance and infodemiology. See his 2009 paper here .
But to Twitter. As the previous poster points out, it is not without its problems
My own quick dive into Google Scholar turned up a 2011 paper about Twitter demographics by Mislove and colleagues (See here.)
Mislove concludes this paper with the statement that “Twitter has received significant research interest lately as a means for understanding, monitoring, and even predicting real-world phenomena. However, most existing work does not address the sampling bias, simply applying machine learning and data mining algorithms without an understanding of the Twitter user population. In this paper, we took a first look at the user population themselves, and examined the population along the axes of geography, gender, and race/ethnicity. Overall, we found that Twitter users significantly overrepresent the densely population regions of the U.S., are predominantly male, and represent a highly nonrandom sample of the overall race/ethnicity distribution.”
The “elite tweet” problem has also been discussed by Wu and colleagues in this 2011 paper from Yahoo Research labs. ( See here .) Wu and colleagues describe the disproportionate influence of a small number of tweeters in the Twitterverse
So what does this mean?
I’d suggest that without understanding who is doing the tweeting, and who is likely to be reading, the anecdotes you’ve supplied about twitter may be a bit misleading.
Don’t get me wrong – it is a great PR channel for promoting one’s products to the media. Simon Chapman’s anecdote is a testament to the potential of effective targetting of an elite message to a very narrow audience.
So, is there a more dynamic space that Twitter for public health and social debate?
Yes. It’s called the internet. With all its blogs, all its news and all its glory.
Just now, I think studying internet search patterns for public health is a lot more interesting.
Underpinning these studies is the principle that each Internet search is a behavioral measure of an issue’s importance to an individual. If individuals are concerned or interested in an issue, they are more likely to search for information related to that issue. The relative importance of an issue can thus be inferred from the volume of search queries for a specific term or terms representing that issue.
There are plenty of papers about the use of internet search for disease surveillance and policy analysis, See for exmaple Ginsberg J, Mohebbi M, Patel R, Brammer L, Smolinski M, Brilliant L. Detecting influenza epidemics using search engine query data. Nature 2009 Feb 19;457(7232):1012-1014, and Reis B, Brownstein J. Measuring the impact of health policies using Internet search patterns: the case of abortion. BMC Public Health 2010;10:514).
Further, there is a whole lot more searching than tweeting going on, and the patterns left by search are actually easily searchable using tools like Google Trends. Across many countries and languages.
(See google.com.au/trends.)
Finally, Google Flutrends really showcases what aggregated search data can do for public health today.
See google.org/flutrends .
Discuss.
Andrey Zheluk
PhD candidate
Menzies Centre fo Health Policy
I agree Andrey; it’s the wider impact and potential of the internet that is most interesting. But I don’t think it’s quite right to frame Twitter as purely a PR channel. It’s so much more interactive and disruptive than that (eg see IndigenousX Twitter account, which I hope to write more on soon). And another link that may be of interest: Canadian efforts to use social media inc Twitter to increase uptake of evidence: http://www.slideshare.net/HealthEvidence/pp23-health-evidence-social-media-final
PS Just to explain my correction (which was made in haste). Initially I cited the number of Google Scholar results when searching on “Twitter” (more than 63,000) and “Twitter and health” (more than 32,000). It turns out that after about page 24 on the first search and page 4 on the second search, the search yields articles that are not about Twitter per se, but have an option for sharing them via Twitter (Yes, I blush at not checking this out more thoroughly first time round and am very grateful that Deborah Lupton was quick to get in touch). So – if you’re searching for scholarly articles on Twitter, beware (and perhaps use multiple databases/types of searches. Other suggestions?
Melissa nothing to blush about re the figures – the whole net measurement field is crazy and based on various flawed numbers.
I’d be curious to know what those who here promote twitter are doing. How many feeds do you subscribe to? how many posts are on each feed daily roughly? How many do you read? How many do you click through?
Why is it better than RSS, newsfeeds, emails lists and specialist blogs, website and aggregate blogs and websites etc?
How many twitts do YOU make a day? Do you make them or some “communications” person in the organisation?
Audrey – great stuff. I think the big news is the old news = The Internet. It’s often crowded out by the latest thing – which almost inevitably has a commercial imperative. Remember when institutions all had to be on MySpace? Then Facebook?………
I just threw out an old box of Epi-Info on floppy disks and a manual from the late 80s. http://en.wikipedia.org/wiki/Epi_Info
I wiped a tear from my eye but what can you do – no-one even knows what a floppy disk is these days… ahh…
Melissa when I asked YOU how you use twitter I didn’t mean you in particular I meant all youse who use twitter.
Andrey – sorry – spell checker
Dr Whom, in case you miss this post, it helps show why Twitter is so much more than a newsfeed:
https://croakey.org/some-thoughts-on-how-the-public-health-sector-can-learn-from-indigenousx-and-its-pitch-to-the-shortyawards/
Having said that, Twitter is my main newsfeed. I need a T-shirt “everything I learn, I learn from Twitter”. I know that is an invitation for some sharp riposte but it’s fairly true. Twitter refers me to books, articles, people, presentations etc that I otherwise may not have come across. More than that, it enables conversations, creativity, connections.
Enough from me…
My T shirt, were I ever to wear one, says, “i think you’ll find
it’s a bit more complicated than that ” and whereby I stand next to someone with a simple slogan t shirt