Tweets and Likes

by Joshua Tucker on January 30, 2012 · 13 comments

in Blogs

Here at The Monkey Cage we allow people to “Tweet” posts to their Twitter followers, and “Like” posts to their Facebook friends. Lately I’ve noticed that some posts get more tweets than likes, some get more likes than tweets, and others get roughly the same amount. Anyone have any idea why? Is there any research on this sort of thing on any other blogs? If anyone wants to scrape the data off of The Monkey Cage and run some analyses, I’d be happy to post the results as a guest post (would make a great paper for a stats class!). In the meantime, I’m just as interested in guesses.

To get us started, I’ll throw out three potential hypotheses:

  1. Humor vs. wonkishness hypothesis: The funnier a post, the more likely it is to go on Facebook; the wonkier the post, the more likely it is to get tweeted.

  2. The graphics hypothesis: The more graphics, the more likely it is to go to Facebook. The more text, the more likely it is to be tweeted.

  3. The source of visitors hypothesis: Visitors outside academia are more likely to post to Facebook; academics who read blogs are more likely to tweet.

To be perfectly clear, I have no idea if the data supports any of these claims. Who else has ideas?

{ 13 comments }

Andreas Moser January 30, 2012 at 3:53 pm

Among my blog posts, those with photos are always the most popular ones by far. For example this recent one about the topless Ukrainian protesters at the World Economic Forum at Davos: http://andreasmoser.wordpress.com/2012/01/28/ukrainian-girls-show-us-how-to-protest/ – They get linked on Facebook a lot, not so much tweeted.

Eric January 30, 2012 at 3:58 pm

A colleague of mine has been working on questions of this sort. His name is Adam Schiffer; I know he did one conference paper examining factors that affected blog posts at the DailyKos. I’ll have to talk to him about his work since then.

Adam Schiffer January 30, 2012 at 4:36 pm

Indeed, I did a study roughly analogous to Dr. Tucker’s suggested analysis, albeit on an earlier iteration of social media: the Daily Kos diaries ecosystem. I accounted for variation in recommendation patterns (the closest thing we had to “likes” and “tweets” in the middle of the last decade) with attributes of diaries such as length, the number of prior diaries by the author, and the quantity and quality of links. The first version appeared as the 2007 APSA paper, “Between Pajamas and Pulitzers . . .”

I agree with the comment below that this sounds like two different analyses: what explains variation in tweets? And what explains variation in likes? While the differences in explanatory variables between the two could be interesting, I think most bloggers (and analysts thereof) would be more interested in why some posts garner external attention while others drop to the bottom of the page with relatively little notice or remark.

Nat January 31, 2012 at 1:56 am

Adam, that article sounds interesting, but I can’t seem to find a copy online. Do you have a link to it?

Adam Schiffer January 31, 2012 at 11:35 am

I always forget that APSA’s archive is gated. Email me at: (a) dot (schiffer) at (tcu) dot (edu).

Peter Grier January 30, 2012 at 4:04 pm

Joshua – as a journalist for whom web hits are a big issue, I think this is a fantastic subject for study. I’ve put some thought to this myself in the past, when I noticed that stories can amass large numbers of page views, with few likes, and vice versa.
My personal hypotehsis is that the “like” button is an instant decision that relates to whether the headline confirms a personal belief.

Gary January 30, 2012 at 4:20 pm

Aren’t we really trying to explain two things at the same time here? What provokes tweets, and what provokes likes?

My hypothesis for the former would be:

* Posts with higher number of tweets are initially tweeted by someone with a lot of followers.

This encompasses hypotheses #2 and #3 I think.

I would test the first part of hypothesis #2 (more graphics = more facebook) independently of twitter effect.

Richard Bridger January 30, 2012 at 5:30 pm

Things that might make a difference (based on my own experience):

The audience effect: my Facebook friends are mostly my friends, my Twitter followers are mostly people with shared (and obviously wonkish) interests. So my wonkish stuff goes on Twitter, the stuff that would be interesting down the pub for my friends who don’t like wonkish stuff goes on Facebook (or both).

The medium itself: Facebook stuff stays on a feed longer – if I post or like 5 things that might be interesting, they can clog someone’s feed for half a day or more. On Twitter, I can post 5 things and they can easily glide over them. So I have to be more discerning with my Facebook links.

The personal endorsement effect: If I like something it suggests I agree with it. But if I just want to flag it as something relevant though not particularly my view, then I can do that more easily with Twitter.

Interesting difference is that you don’t have a share function for Facebook. If I want my Facebook friends to see something then I tend to actively share it on my account (I think that makes it more prominent, though I may be wrong). That might grab you some more granularity if it were possible to track.

David Benson January 30, 2012 at 6:16 pm

I have a hypothesis and a theory:
Hypothesis: Posts which liberals want to distribute are more likely to get liked, whereas conservatives wish to distribute are more likely to get tweeted (two complementary hypotheses, I know).

Theory:
Facebook became popular during the Bush presidency and Republican Majority, and therefore liberals found that as an easy place to organize while “in the wilderness”, and it remains so. On the other hand, twitter didn’t become popular until around the time that Republicans lost the majority, and eventually the presidency, therefore conservatives tended to organize there while “in the wilderness”. (See this graphic for some mild support. I also remembering a study that claimed this distribution empirically about two years ago, but I couldn’t find it. Twitter and facebook remain common — contrast with facebook and myspace — and therefore there has been no reason to change, but if facebook or twitter were to begin to lose market-share, then you might see a change where as one of the two captures the market, the party that is affiliate with the loser begins to shift to the winner.

Potential test:
Follow trackbacks to websites which are identifiably partisan for your correlation (obviously, this cannot include facebook and twitter), and test the correlation.

Joanna Bryson January 31, 2012 at 4:12 am

Do you use google analytics? I just noticed that a blog post on open access was accessed almost significantly more from google+ than twitter, despite my having very significantly more followers on twitter, and not at all from Facebook, despite having the same order of magnitude of “friends” as twitter followers. I suspect twitter has more people in too much of a hurry to click links than g+, which attracts not only geeks but people interested in public discussion.

Speaking of public, I favor (also without data) your theory 1. Fb is for entertaining friends & family, but twitter is a public display of fitness & competition for bandwidth.

Mike Jensen January 31, 2012 at 4:37 am

I think Bridger is on to something. Likes and tweets are communicative actions and when we communicate we do so with an eye to the audience. Many academics have friended their family members and friends — perhaps students as well — who are on Facebook for social rather than professional reasons. Hence, one may be more inclined to select information that their non wonk/academic/political geek audience would enjoy. It is probably not an accident that this relationship obtains between Facebook and Twitter, but suspect it has more to do with the architecture of the platforms rather than US domestic politics. Given differences in organization of Twitter and Facebook (albeit less pronounced when accessed on phones and tablets), there is an organization of the digital you on Facebook, reinforced by the new timeline feature. Whereas the central feature of Twitter is the engagement with a short text, Facebook situates you in a web of family and friendships, relationships statuses and photos. Rather than study the attitudes of those rediffusing the blog, one might study the content of a their Twitter and Facebook posts along with the networks of friends and followers.

JP January 31, 2012 at 11:24 am

What’s the st. dev and mean of the number of tweets/post and likewise likes/post?

Vit February 3, 2012 at 3:55 am

I’m Czech and I share your posts on FB because Europe is still behind US in terms of Twitter popularity. So if you do have a lot of traffic from Europe, do not forget to take this ainto account.

Comments on this entry are closed.

Previous post:

Next post: