Social Data Signals Y1, W3

For those who didn’t have a chance to sign up, we thought we’d pop our third issue of Social Data Signals onto the blog. If you like what you see make sure to subscribe by following this link.

Welcome back! Stacks of Pew Internet research this week on social media and teenagers, ethnicity, and news event hashtags. Plus Pinterest pinnability from both their data science team and new research by Curalate.

Give us your feedback on Twitter – we’re @hautepop and @abc3d.

Fran and Jay

Understanding Audiences

From Twitter to Instagram, a different #Ferguson conversation [Pew Research Centre]
Hashtags provide a point of entry into online conversations on major news events – but the way they operate varies sharply by platform.

Is politics on Twitter a man’s world? [BBC]
Spoiler: apparently yes (though a test Jay did suggests that the Twitter gender algorithm over-counts men. So take with a pinch of salt)

Social media preferences vary by race and ethnicity [Pew Research Centre]
Whereas Facebook is evenly spread across all audiences, Pinterest is the “whitest” platform. NB age explains this as much as race.

Measuring Online Social Bubbles [arXiv]
The filter bubble problem, quantified, for the first time.

Network archaeologists discover two types of social network growth in historical Facebook data [Technology Review]
How the age of a community shapes the way the network grows over time. (“Network archaeologist” also an exciting new job title)

Visual Social Media

The hidden biases of Internet memes [Washington Post]
The racial and gender biases lurking behind the top 50 English language memes.

Overcoming an eating disorder on Instagram [The Atlantic]
The unexpected consequences (and insight) of posting food pictures on Instagram.

ShareThis study looks at Pinterest and Purchases [Web Pro News]
Why Pinterest is a sales conversion machine, who’s behind it and what do people buy the most?

The maker of the Trollface meme is counting his money [Kotaku]
How a modern meme was born, and how it went from teenage bedroom to copyright and global fame.

Social Platforms

Teens, Social Media & Technology Overview 2015 [Pew Internet]
Aka talk of Facebook’s death among teenagers would seem premature. It’s still the #1 platform for 41%, twice the Instagram rate.

Snapchat’s Non-Vanishing Message: You Can Trust Us [Medium]
“A rare interview with Snapchat tech executives about a transparency report, a bug bounty program, a third-party app ban…and apologies”

Modeling Mediums Of Communication [TechCrunch]
CEO of the Meerkat video streaming app talks media theory, quotes psychoanalyst Jacques Lacan (!), shares useful diagrams.

8 Pinterest statistics that marketers can’t ignore [Sprout Social]
A cheap headline but good stuff in here rounding up research from Curalate, Shopify & Comscore.

This Is the Perfect Pinterest Picture, According to Science [Wired]
More detail on Curalate’s research on what makes pictures pinnable.

Value for data

Studies in social data: how industry uses social media for communications and research [LSE]
Farida Vis reports back from Twitter’s Studies In Social Data meetup

Six burning questions for firms looking to make money from big data [Telegraph]
Professor Andy Neely discusses how firms can capitalise on the use of big data to improve their business model.

What to do when people draw different conclusions from the same data [HBR]
Answer: throw 60 data scientists at the analysis and triangulate results. Everyone’s got 60 data scientists spare, right?

Social Data Science

Twitter Sentiment Visualizer app [NCSU]
Enter a keyword and this tool brings back tweets and visualises sentiment on more interesting emotional models than just positive/negative. Super-easy to use, a great tool for students.

Network structures and dynamics in online social systems[O’Reilly Radar]
Understanding information cascades, viral content and significant relationships.

If algorithms know all, how much should humans help?
[New York Times]
As the stakes rise for the application of data science beyond digital products, marketing and advertising, the accountability of algorithms become more and more crucial.

Data APIs, Design and Visual Storytelling [O’Reilly Radar]
How data APIs are allowing us to focus on good storytelling.

Re: Our Relationship [The Atlantic]
A data scientist tries to quantify love, studying a year of emails she exchanged with her boyfriend.

If you’ve enjoyed this third issue and want to receive weekly social data updates straight to your inbox, please subscribe by following this link.

Also, if you have any feedback, suggestions or articles you’d like to see in the newsletter, please contact Fran and Jess by using the hashtag #smartdata on Twitter. 


Ed Hawes