Modern Metrix

Free Twitter Data!

I am always on the look out for free data to play with, and today stumbled upon awesome depository of Twitter data on Infochimps website. Check it out!

 

Klout vs. Peer Index Scores: analyzing differences

While searching for some articles on influence, I stumbled upon this blog post by Christopher Penn, where he was challenging readers to dig into a small dataset of Klout and PeerIndex score to explore the differences between these two metrics of influence. I quickly ran some numbers which I included in my comment to his post, but would like also to share here.

The dataset that Chris was kind enough to post contains Klout and PeerIndex scores for 15,737 Twitter users who tweeted at least once in the past month, presumably on topic related to #Marketing (data is available on his blog, in CSV format). I first computed straightforward Pearson's correlation between two scores, which turned out to be .45. The rank-order correlation (correlation between the ranks of the scores instead of their actual values) was slightly lower (.43). Finally, I assigned the scores to quintiles (5 equal groups; think of them as top 20% of users with highest scors, then next top 20%, etc., ending with bottom 20%) and looked at the correlation and cross-tabs between the quintiles. The correlation was .42, the average overlap between two scores (that is, user who was in the top 20% on Klout score wa also in the top 20% on PI score was about 30%). And here is a modified confusion matrix for all 5 quintiles, showing the overlaps between the groups:

As you can see, the top and bottom 20% had higher overlaps. Overall, there is a positive relationship of moderate strength between two metrics. But only moderate.  One possible explanation for discrepancies is that Klout and PeerIndex use different reference groups. Remember, the scores are computed, normalized and compared within a larger group of users. Since it is nearly impossible to download and process data on ALL Twitter users, so most likely, these services draw random samples of people on Twitter and then use them as reference groups. 

4.74 Degrees of Separation

Interesting article in NYTimes about most recent replication of Milgram's famous "small world experiment". The original research introduced concept "six degrees of separation" - that any two people in the world are separated by only six intermediate connections. According to Facebook scientists, who analyzed data on 721 million active Facebook users with total of 69 billion friendships, that number now is down to 4.74. Of course, I would question definition of "friendship" used in the study (it was defined simply as a being "friends" on Facebook), nevertheless, the methodology and analysis are interesting. 

More detailed report is available on Facebook Data page here.  

Free Book on Social Media Measurement and Analysis

Great FREE book on Social Media Measurement and Analysis from Radian6  http://bit.ly/cm8pS1

What news articles are more likely to be shared?

This interesting study by researchers from University of Pennsylvania looked at how the topic, content, lenght, style and other characteristics of the news articles affect the likelihood of people sharing them.

"To make sense of these trends in “virality,” the Penn researchers tracked more than 7,500 articles published from August 2008 to February 2009. They assessed each article’s popularity after controlling for factors like the time of day it was published online, the section in which it appeared and how much promotion it received on the Web home page.

A random sample of 3,000 of these articles was rated by independent readers for qualities like providing practical value or being surprising. The researchers also used computer algorithms to track the ratio of emotional words in an article and to assess the relative positivity or negativity..

..More emotional stories were more likely to be e-mailed, the researchers found, and positive articles were shared more than negative ones. Longer articles generally did better than shorter articles, although Dr. Berger said that might just be because the longer articles were about more engaging topics. (The best way to test that, he said, would be for The Times to run shorter and longer versions of the same article that would be seen by different readers.)

..Surprising articles, like one about free-range chickens on the streets of New York, were also more likely to be e-mailed — which was a hardly a surprising discovery, of course. But the researchers also kept finding popular articles with a quality that went beyond surprise."

Grab some free data on 215 mln. Facebook users, for free!

Pete Warden is making data on over 200 mln Facebook users available to academic research community this week. The fields in the dataset are limited to publicly available info such names, location, fan pages, and friends lists. Pete also shares some preliminary insights:

"In almost every state in the Southern U.S., God is number one most popular Fan page among Facebook users. Among people in the L.A., San Francisco and Nevada regions? "God hardly makes an appearance on the fan pages, but sports aren't that popular either," Warden writes. "Michael Jackson is a particular favorite, and San Francisco puts Barack Obama in the top spot." In the Oregon and Idaho region? Starbucks is number one.

In the Mormon-influenced areas of Utah and Eastern Idaho, the most popular Fan pages are The Book of Mormon, Glen Beck and the vampire book Twilight, which was authored by a Mormon.

The bulk of Warden's posted analysis yesterday was about location networks. People in the western U.S. tend to have Facebook friends all over the country; people in the southern U.S. tend to mostly be friends with people who have remained in the same area."

How local is the social networking?

This study from Hebrew University suggests that geographical distance between the members of online social network and intensity of their communications are inversely related. They studied the messaging habits of 100,000 Facebook users by zip code and found that the closer are the users in a physical space, the higher is density of their e-mail communications.

Purchase Influence in Social Networks: A Study from Harvard Business School

Interesting study "Do Friends Influence Purchases in a Social Network" from Harvard Business School (PDF). Excerpt from the abstract:

"Our results show that there are three distinct groups of users with very different behavior.
The low-status group (48% of users) are not well connected, show limited interaction
with other members and are unaffected by social pressure. The middle-status group (40%
users) is moderately connected, show reasonable non-purchase activity on the site and have a strong and positive effect due to friends’ purchases. In other words, this group
exhibits “keeping up with the Joneses” behavior. On average, their revenue increases by 5% due to this social influence. The high-status group (12% users) is well connected and
very active on the site, and shows a significant negative effect due to friends’ purchases.
In other words, this group differentiates itself from others by lowering their purchase and
strongly pursuing non-purchase related activities. This social influence leads to almost 14% drop in the revenue of this group. We discuss the theoretical and managerial implications
of our results."

 

TV Trends search engine!

Now, this is amazing: TV Trends let you search mentions of any word or phrase on TV. You get a plot of dates when that word was mentioned, genre of the TV program, station, AND transcript of what was said.

Twitter Sentiment Analyzers: Review

Chris Near of KDPaine & Partners reviews 5 tools for analyzing sentiment of posts on Twitter in the lastest issues of "The Measurement Standard".