Facebook’s Valentine’s Day gift to all of us: data about our relationships

You could learn about society from official sources like the U.S. Census. Or Facebook can help you.

In the run-up to Valentine’s Day this week, the social-networking powerhouse released lots of nifty observations about people’s relationships, gleaned from the untold number of petabytes of data the company retains about its more than billion-strong user base.

[aditude-amp id="flyingcarpet" targeting='{"env":"staging","page_type":"article","post_id":904632,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"big-data,business,","session":"C"}']

The discoveries from Facebook data scientists hit on couples’ ages and religious affiliations, as well as the places where singles enter relationships, and the places users check in during the week of Valentine’s Day.

It isn’t the first time Facebook’ data scientists are sharing aggregated observations with the public. They’ve previously brought data to questions like what makes people start rooting for football teams, when did people start selecting the Human Rights Campaign logo for their profile images, and how many Facebook friendships exist between parents and children.

Earlier this month, Twitter got into the game, too announcing that it would give “a handful of research institutions” access to its corpus of public tweets. The trend here is that social networks are doing more to promote their data as a product for external consumption.

This week, we’re getting a whole package from Facebook pertaining to people’s relationship status — one of the most significant aspects of the connections among users. It’s the kind of thing you might expect from a group like the Pew Research Center, based on a survey. But no — it’s from Facebook. And the findings are pretty interesting.

Facebook data scientist Carlos Diuk studied timeline posts going between Facebook users who started relationships. He found that in the 100 days leading up the day when users change their relationship status, the number of timeline posts generally goes up. The high point comes 12 days before the beginning of the relationship change, with 1.67 posts a day. After that, timeline post activity drops off.

“Presumably, couples decide to spend more time together, courtship is off, and online interactions give way to more interactions in the physical world,” Diuk wrote in a blog post yesterday.

Diuk analyzed how the sentiment of the timeline posts changes over time, and he found positive feelings increasing significantly right around the day when the relationship begins.

Mike Develin, another member of Facebook’s data science team, figured out which cities are most popular among single people. Detroit, Los Angeles, New York, Miami, and Memphis are the top five.

[aditude-amp id="medium1" targeting='{"env":"staging","page_type":"article","post_id":904632,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"big-data,business,","session":"C"}']

Develin also identified the five cities with the highest probability of a relationship being formed: Colorado Springs, Colo.; El Paso, Texas; Louisville, Ky.; Fort Worth, Texas; and San Antonio, Texas.

Notice the lack of overlap for those 10 cities.

“In a city where everyone is paired up, the incentive to pair up is even stronger, while cities like New York and Miami are places that people go to be single,” Develin wrote today in a blog post of his own.

Earlier this week Develin took a look at countries and relationships. He brought religion into the equation, too. Interfaith relationships were most common in Spain in a sample of users, he wrote in a post about the research.

[aditude-amp id="medium2" targeting='{"env":"staging","page_type":"article","post_id":904632,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"big-data,business,","session":"C"}']

Romania and Portugal stand out as places for being popular places for intermarriage. Christian Orthodox rules in the former and Catholicism is the leading religion in the latter, Develin wrote.

In the U.S., younger people are more likely to end up in interfaith marriages than older people, according to Develin’s findings.

Develin explored age gaps as well. In the U.S., he noticed that as people get older, the average age gap rises.

“This makes sense, since younger couples are more likely to have met in school,” he wrote in a blog post about the connection between age and relationships. He restricted the evaluation to people in relationships whose average age fell between 20 and 49.

[aditude-amp id="medium3" targeting='{"env":"staging","page_type":"article","post_id":904632,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"big-data,business,","session":"C"}']

On an international level, Develin observed that in opposite-sex relationships, the male is older than the female 67 percent of the time. It’s the other way around 20 percent of the time.

It occurred to Develin that wealth could play a role in age gap.

“There appears to be some connection to affluence level of the country, perhaps because in countries with higher educational expectations, couples are more likely to meet in college,” he wrote.

Facebook data scientist Bogdan State found some noteworthy tidbits while inspecting the duration of relationships. For one thing, the longer the relationship, the less of a chance of a breakup, he wrote in a blog post.

[aditude-amp id="medium4" targeting='{"env":"staging","page_type":"article","post_id":904632,"post_type":"story","post_chan":"none","tags":null,"ai":false,"category":"none","all_categories":"big-data,business,","session":"C"}']

State pruned the data down to a narrow selection: U.S. users who initially chose a relationship status other than “married” and were at least 23 when they entered relationships that lasted at least three months on Facebook. He then only used those relationships that started between January 2008 and December 2011.

For that subset of relationships, State found that breakups tended to hit a high point in May through July, while in February 2009 and February 2010, breakups decreased. Maybe it’s because of Valentine’s Day, State wrote. Then again, he wrote, “could it be that a rebounding U.S. economy played a role?”

Sure, these deep dives gives us plenty to chew on about our relationships. But they also hint at how much one could learn by getting a chance to tap a vast data pool like Facebook.