It’s increasingly rare (at least from a digitally entrenched perspective) to imagine a journalist watching a presidential debate without simultaneously watching his or her tweets. This is certainly fine, and in many cases, helpful. But with CJR’s recent piece on “pack journalism” and in light of some recent studies on Twitter makeup and preferences, I figured it’d be good to review a handful of the findings together and what they may mean for journalists.
The larger aim is that a thorough understanding of the Twitter community – placed at least in the back of one’s head – could help one from being heavily influenced by that scary hive-mind (if it’s true), and regardless, put into perspective the general sentiments that may soak in when one repeatedly scans TweetDeck.
Understanding the community in any medium you regularly use, not just Twitter, is a good practice. There is always a filter bubble wherever we engage online—we tend to regularly admit that, and some of us take steps to pop it by whom we follow and what we search for. The recent findings I’ve compiled about Twitter, however, seem of a particular importance, for they shed some light on what may be a wider filter bubble (“filter fish tank”?) of what is increasingly many journalists’ anchor.
The items in the outline below are largely intended as 1) gut-checks for journalists while particularly covering the X number of tweets sent during a debate, sentiment during a large news event, and other social-focused stories, and 2) things to keep in mind just for day-to-day Twitter use as your resource (or “newspaper”). In short, if Twitter somehow influences your reporting and coverage, it’s good to have a handle of its user and usage realities before any influence on your decision-making spreads outward to other media and consequently, the general public’s knowledge. (And, because knowledge influences it, the general public’s opinion.)
A line from a BBC piece on a connected topic (“Is Twitter good for democracy?”) captures the idea: “even voters who aren’t on Twitter find themselves influenced by the Twitter narratives.” It’s worth knowing who is on Twitter, because who is there (and active) does have some definite effect, whatever it may be. Here is just my look and half-formed thoughts—I encourage you to look at recent data and do your own analysis.
1) 13 percent of Americans “ever use Twitter or read Twitter messages”
“While news gathering is very common among Twitter users, the overall reach is limited because the audience remains relatively small.”
Twitter may be an increasingly popular and influential stronghold for journalists in the U.S., but it’s worth noting that according to Pew’s data, it isn’t hugely popular among the public. While some early 2012 data suggests that 15 percent of Americans are “on” Twitter, that may just mean they have an account—according to this more recent data, just 13 percent of Americans “ever use Twitter or read Twitter messages.” Comparatively, more than half of Americans (54 percent) use other social networking sites (such as Facebook, Google Plus, or LinkedIn). As such, “far fewer” people get news on Twitter than elsewhere on the social web, even though as a journalist, you couldn’t imagine not using it for news.
An according to that same data set, just 11 percent of Americans “ever” see news on Twitter, and only “3 percent got news there yesterday.”
So, if the data holds up, while news gathering is a common practice on Twitter, the folks doing the gathering and sharing – and using – are a small portion of the American public. The reach is small, and consequently, so is any sample of opinion (or sentiment).
Reminder from the stat: The voices aren’t everyone.
2) Roughly half of all tweets are sent from users in the U.S.
Out of the 36 million Twitter users that social media marketing firm Beevolve analyzed in its study published last week, more than half – about 51 percent – were from the U.S.
Combining that data with Pew’s data about Twitter usage in the U.S. means that about 13 percent of Americans make up about half of the world’s Twitter users.
For perspective, the U.K was second in the worldwide distribution of Twitter users, making up roughly 17 percent of the service’s user base. After that, country representation drops significantly in terms of percent: Australia has about 4 percent, Brazil is fourth with about 3.5 percent, and Canada is fifth with just under 3 percent. Iran, ninth on the list of highest number of Twitter users, has 0.88 percent of Twitter’s user base.
So a small portion of Americans are “in control” of much of what’s on Twitter.
Reminders from the stats thus far: The voices aren’t everyone, and they have a large control over what’s available on all of Twitter.
3) Democrats are more likely than Republicans or independents to use and place value in social networking sites
“In each activity, Democrats who use social networking sites are more likely than Republicans or independents to say the sites are important.”
From the Pew report, here are those activities. The percentages who agreed with the specified point at all are listed first, followed by the political breakdown
36 percent of social networking site (SNS) users say the sites are “very important” or “somewhat important” to them in keeping up with political news.
Of those surveyed, 48 percent of Democrats say the sites are “very important” or “somewhat important” to them for this. That compares to only 34 percent of Republicans and 33 percent of those grouped as Independent.
26 percent of SNS users say the sites are “very important” or “somewhat important” to them in recruiting people to get involved in political issues that matter to them.
Thirty-five percent of Democrats say the sites are “very important” or “somewhat important” to them for this. That compares to only 25 percent of Republicans and 22 percent of Independents.
25 percent of SNS users say the sites are “very important” or “somewhat important” to them in finding other people who share their views about important political issues.
Thirty-four percent of Democrats say the sites are “very important” or “somewhat important” to them for this. That compares to only 23 percent of Republicans and 22 percent of Independents.
25 percent of SNS users say the sites are “very important” or “somewhat important” to them for debating or discussing political issues with others.
Thirty-two percent of Democrats say the sites are “very important” or “somewhat important” to them for this. That compares to only 24 percent of Republicans and 23 percent of Independents.
You’ll note that these stats are for “social networking sites,” not just Twitter. But as we discussed in the first listed statistic, many more Americans use social networking sites that are “other than Twitter.”
We don’t know this data specific to Twitter users. (Yet. Hopefully it’s on the way.) But note again that the Twitter user-base (note not even “those tweeting”) is just about 13 percent. It’s important to pair this with data on political affiliation and values in general—not all political parties value social media the same. You can probably extrapolate that to not all opinions value social media the same—which may seem like an understood, but it’s important to be clear about when analyzing any social media for stories (and in the stories themselves).
Reminders from the stat thus far: The voices aren’t everyone, and they have a large control over what’s available on all of Twitter. Politically, the voices vary in what opinions they choose to express and the value they place in doing so.
4) 12 percent of Twitter users have protected accounts on average
Source: Beevolve’s study
About 12 percent of Twitter users worldwide have protected accounts—meaning their tweets don’t show up in streams, unless you’ve requested and been given access. There isn’t data from Beevolve about what the percentage of protected accounts are in the U.S. (and I can’t find this elsewhere). But being that the U.S. has the most users, it seems reasonable to guess that the majority of the protected accounts could also from the U.S.
That’s a variable of Twitter sentiment, opinion, reaction and so on that’s not in play during most analysis. It’s worth considering what these accounts may be saying. At the minimum, it’s worth noting that they’re there.
Reminders from the stats thus far: The voices aren’t everyone, and they have a large control over what’s available on all of Twitter. Politically, the voices vary in what opinions they choose to express and the value they place in doing so. Who knows what some of them are saying, because you can’t see them.
5) There are 6 percent more female users on Twitter, and they tend to tweet more (More demographic stats are also worth considering)
Source: Beevolve’s study
When Beevolve’s study came out, the first headline I saw was at GigaOM: “Typical Twitter User is a Young Woman with an iPhone and 208 Followers.” Many pieces joked about or brought up stereotypes and what the stats had to say. It’s an idea worth engaging. Data suggests that despite how much weight you give it as a journalist (or media company), there are indeed certain demographics that more actively use Twitter.
Some of the demographic data in Beevolve’s study is admittedly skewed because it goes off of self-disclosed profile info. More younger people are likely comfortable disclosing their age, for instance, so it makes sense that you would see the data of self-disclosed age showing that 74 percent of Twitter users are between the ages of 15 to 25. Only 0.45 percent have disclosed their age on Twitter, so that sample is likely too tiny to make many age assumptions on.
In the case of gender anaylsis, that’s probably more trustworthy. Beevolve’s study determines its gender stats off a combination of self-disclosed profile info, names, and profile pictures.
It seems entirely possible that the demographics who use Twitter aren’t representative of the entire American public either (or the world, for that matter).
Reminders from the stats, in sum: The voices aren’t everyone, and they have a large control over what’s available on all of Twitter. Politically, the voices vary in what opinions they choose to express and the value they place in doing so. Who knows what some of them are saying, because you can’t see them. More user demographics including gender and age are also worth noting when you’re looking at Twitter’s sentiment, tweets per minute, reactions and so on.
In the lower case “d” sense of the word, what’s on Twitter isn’t universally democratic. Be mindful of this in any web real estate or TV time you give to what’s on Twitter, and how much you use it as a resource yourself.
Like everywhere else in journalism and writing, know your audience. And in Twitter’s own unique two-way, too, know your source.
- Friday Roundup: The Week in Journalism
- New Draft of SPJ's Ethics Code Now Available
- ClearVoice Measures and Scores Writers' Social Influence: How Do You Rank?
- Center for Investigative Reporting to Launch Public Radio Show