Now that I’ve got some of my database woes solved I’m digging back into some interesting Twitter data again. Today, we’ll take a look at the set time zone of active Twitter users. As in the last analysis, this one is based on active Twitter users. Specifically I’m looking at the results from a set of almost 3.5 million recently active Twitter users.
When looking at Twitter users, it’s hard to use location information from profiles since some people may use a specific location (e.g., Atlanta, GA), others will use latitude and longitude coordinates, and still others will use things so arbitrary that they’re unhelpful (e.g., Earth, the Internet, etc.). Unfortunately the time zone specification also falls prey to this same issue since supplying the information is optional but the benefit is that the data, if available, is a simple number that can be easily manipulated and collated to provide meaningful information.
So here is a chart of the 3.5 million users I had captured when I ran this analysis.
|Japan+%2F+Korea+(16.74%25)|Eastern+USA+(10.59%25)|Western+USA+(8.64%25)|Central+USA+(7.85%25)|Brazil+%2F+Argentina+(7.78%25)|Brazil+%2F+Chile+%2F+Caribbean+(4.03%25)|Hawaii+(3.05%25)|Europe+%2F+Africa+(UTC%2B1)+(3.66%25)|Russia+%2F+Thailand+(3.05%25)|Mountain+USA+(1.97%25)|China+%2F+Australia+(1.97%25)|Alaska+(1.37%25)|Venezuela+(0.89%25)|Europe+%2F+Africa+(UTC%2B2)+(0.71%25)|Russia+%2F+eastern+Australia+(0.70%25)|India+%2F+Sri+Lanka+(0.47%25)|Africa+(UTC%2B3)+(0.37%25)|others+(0.67%25))
Of the data that I had available, about a quarter of users had no specified time zone. I’m assuming these users are uniformly spread throughout the world so I’m just going to ignore this group.
The first interesting insight is how much larger the east coast (of the USA) users are than west coast (again, of the USA) users. I would have expected that the west coast would have been a larger number of users just because 1) Twitter is based in the Valley and 2) there are a lot of people in California. But as the data shows, this isn’t true. This same analysis has repeated over different time frames over the last week or so as I created and destroyed my database. I’m not sure what this means but it might mean that Twitter (and others) should be paying as much, if not a little bit more, attention to the east coast as has been paid to the west coast.
The other data points that pop up (but are hidden on the chart above) relate to two of the smallest time zones. The second to the last smallest is at UTC+4.5. This time zone is solely for Afghanistan. I would have thought that Afghanistan would have had the smallest representation due to challenges with Internet access and overall security. But the smallest time zone was indeed a surprise, at least for me.
Myanmar (formerly Burma) and the Cocos Islands are covered by the smallest time zone, UTC+6.5. I would have assumed there would be no Twitter users in Myanmar given that the military junta has blocked Internet access in the past and continues to restrict free and open access. I spot checked a few of these accounts and they seem to by Myanmar related. So either these folks are nearby Myanmar, mistakenly set their time zone to UTC+6.5, or somehow they have acquired Internet access within the country. Either way, I’m declining to share the numbers or specific accounts related to this area for obvious reasons. Personally I hope these people are in country and if so, I find it encouraging that they’re using Twitter to discuss their issues and the facts on the ground in Myanmar.
More to come as I continue looking at this data.