Showing posts with label stats. Show all posts
Showing posts with label stats. Show all posts

Sunday, July 17, 2011

Is Google+ Becoming More Female?

by 

If you believe the “unofficial statistician” for Google+, the new social network isn’t nearly as male-dominated as previously reported.
Paul Allen, the founder of Ancestry.com and the man who released a report estimating that Google+ was about to reach 10 million users, says that the male-to-female ratio on Google+ is not as disparate as some reports, including those published by Mashable, have seemed. Instead, Allen’s latest figures show that 33% of Google+ users, as of July 14, 2011 are female.
To be clear, this still means that Google+ is still a male-dominated network (at least for now), but the breakdown isn’t as extreme as the 90% male, 10% female stat from SocialStatistics or the 75% male, 25% female stats from FindPeopleonPlus.
Allen’s methodology in estimating Google+ statistics varies from some other Google+ data sources.
This is how Mashable’s Stan Schroeder described the methodology earlier this week,
He sampled a number of surnames from the U.S. Census Bureau data and compared it to surnames of Google+ users. By comparing surname popularity in the U.S. with the number of users on Google+ with each surname, he can guesstimate the percentage of the U.S. population that signed up for Google+. Finally, he calculated a ratio of U.S. to non-U.S. users to generate an estimate for the number of Google+ users worldwide.
We’re not sure how census data can be applied to gender, especially on a social network, but we’ve reached out to Allen for additional information on how his data was composited.
Still, Allen’s figures show that from July 4, 2011 through July 14, 2011, the male/female gap closed considerably. On July 4, 2011, Allen estimated that 23% of users were female. By July 14, 2011, that percentage was up to 33.6%.
In its own videos and marketing for the service, Google put a lot of emphasis on women, which to us, implies that women are a prime target for Google+.
In our own unscientific estimates, we’ve seen more women joining Google+ over the last few days than when the service first launched.
That leads us to our question — do you see more women joining Google+ in your own circles? What do you think it will take for the gender ratio to start to balance out? Let us know in the comments.

Tuesday, February 15, 2011

Where Twitter Trending Topics Really Come From [STATS]

o by Ben Parr
HP‘s Social Computing Research Group has released the results of a new study that dives into the anatomy of Twitter’s Trending Topics.
For its research, HP analyzed 16.32 million tweets on 3,361 different trending topics between September and October 2010. To get its data, it queried Twitter’s search API every 20 minutes.
HP discovered that Twitter’s Trending Topics algorithm cares more about the specific subject and reach of a tweet than who tweets it or how often it’s tweeted. Around 31% of trending topics are retweets. More importantly, 72% of those retweets come from mainstream media outlet like @cnnbrk or @nytimes. The Telegraph, ESPN, @breakingnews and The Huffington Post all made the list of top retweeted users in at least 50 different trending topics.
“What proves to be more important in determining trends is the retweets by other users, which is more related to the content that is being shared than the attributes of the users,” HP concludes in its research report. “Furthermore, we found that the content that trended was largely news from traditional media sources, which are then amplified by repeated retweets on Twitter to generate trends.”
Tweets from “influencers” have little effect on trending topics. Instead, trending topics often come from news stories tweeted from major news outlets. HP concludes that traditional media still starts the conversation around the most-discussed topics in social media, not the other way around.
HP’s researchers (led by HP Senior Fellow Bernardo Huberman) also analyzed the factors impacting the length of a trending topic on Twitter. It found that the very few trending topics stay at the top longer than 40 minutes. “We showed that the distribution of long-time trends is predictable, as is as the total number of tweets and their growth over time,” Huberman said in the company’s blog post.