Maybe an AI engine has the answer on whether your relationship will end up being a happy one.
In a new study published in the Proceedings of the National Academy of Science, scientists used machine learning to analyze what factors were important in predicting a couple’s relationship satisfaction.
Researchers, led by Samantha Joel and Paul Eastwick, conducted a machine-learning analysis of data from around 11,000 couples and found that relationship-specific characteristics were more important predictors of relationship quality than individual ones.
“Relationships-specific variables were about two to three times as predictive as individual differences, which I think would fit many people’s intuitions,” said Joel, a psychologist from Western University in Canada.
“But the surprising part is that once you have all the relationship-specific data in hand, the individual differences fade into the background.”
Using machine learning to sort through relationship data
The field of relationship science has actually existed for decades and has prompted a significant amount of psychological theory about what makes for happy and unhappy couples.
However, one of the biggest challenges for this growing field is bringing data together on a larger scale. Doing so is needed to bolster the findings of smaller, standalone studies. Meanwhile, conducting large studies can be prohibitively expensive, as well as time-consuming to conduct, in terms of recruiting and interviewing participants.
Machine learning algorithms provide a unique solution to this problem.
Machine learning algorithms are a form of artificial intelligence (AI) that teaches computers to think similarly to how people do: by learning and improving upon past experiences. These algorithms work by exploring data and identifying patterns while requiring only minimal human intervention.
Using this, almost any task that can be completed with a data-defined set of rules or patterns can be automated. This includes sifting through data from various studies, which is exactly what Joel and Eastwick did.
Machine learning confirms that the relationship matters more than the individual
For the study, a team of 86 researchers shared 43 datasets involving over 11,196 couples with Joel and Eastwick. The two then fed this data to a machine learning system called Random Forests, which can test the predictive power of a large number of variables fed to it.
Using this technique, the algorithm determined what kind of reported variables seemed to matter the most in terms of predicting relationship quality
According to Joel and Eastwick, relationship-specific predictors such as “appreciation,” “perceived partner commitment” and “sexual satisfaction” account for nearly half of the variances in relationship quality.
On the other hand, individual characteristics – the ones that describe a partner rather than a relationship – explained 21 percent of the variance in relationship quality. Among these characteristics, the ones with the strongest predictive power for relationship quality are “satisfaction with life,” “depression,” “negative affect,” “avoidant attachment” and “anxious attachment.”
“‘Who I am’ doesn’t really matter once I know ‘who I am when I am with you,’” Eastwick added.
In addition, Joel noted that one partner’s individual difference predictors, such as agreeableness, depression and satisfaction, explained only 5% of variance in the other partner’s relationship satisfaction.
“In other words, relationship satisfaction is not well-explained by your partner’s own self-reported characteristics,” Joel said.
Based on their data, the researchers conclude that things such as negative affect, depression, or insecure attachment are relationship risk factors; however, these matter very little if the couple manages to establish a relationship characterized by appreciation, sexual satisfaction and a lack of conflict.
While the data offers a huge range of potential statistical insights, Joel says that the findings, at their heart, boil down to a pretty simple truth. (Read: Scientists have taught an AI how to smell things)
“Really, it suggests that the person we choose is not nearly as important as the relationship we build,” he said in an interview with Inverse. “The dynamic that you build with someone—the shared norms, the in-jokes, the shared experiences—is so much more than the separate individuals who make up that relationship.”