This Algorithm Can Predict Relationship Trouble

By analyzing the vocal patterns of couples in therapy, an algorithm was able to predict whether a relationship would get worse or improve. Erika Beras reports.

Illustration of a Bohr atom model spinning around the words Science Quickly with various science and medicine related icons around the text

Join Our Community of Science Lovers!

Watch your tone—because it turns out it really isn’t what you say—it's how you say it. At least when it comes to couples in couples counseling. 

That’s according to a study in Proceedings of Interspeech. [Md Nasir et al., Still Together?: The Role of Acoustic Features in Predicting Marital Outcome

Researchers developed a computer algorithm to gauge relationships between spouses based on their vocal patterns. Working with hundreds of recorded conversations from marriage therapy sessions collected over two years, the algorithm was able to predict whether a relationship was going to get better or worse with an accuracy of just under eighty percent. 


On supporting science journalism

If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.


How they did it? The recordings were divided by acoustic features that used speech processing techniques to track pitch and voice warble and intensity. 

These clips from the researcher’s training video illustrate psychological states that characterize distressed relationships. This one, for example, shows “negative affect” and “reactivity” – behaviors that relationship experts believe are troublesome. 

Female: And I want you to just come home at a more reasonable time rather than you know walking in the door at 11. Male: I just don’t think you understand just how much I have to do, what my work entails. Female: Well, what is there to understand? 

The counseling sessions were also tested against behavioral analyses with codes for positives such as “acceptance” and the negatives such as “blame.” Using only that more standardized research method wasn’t as predictable as listening to the vocal expressions.  

Now, these examples are negative as the researchers focused on distressed relationship dynamics. One could imagine the algorithms may also work the same way when looking at positive vocal patterns. Because even married couples sometimes say nice things to each other. 

—Erika Beras

[The above text is a transcript of this podcast.]

It’s Time to Stand Up for Science

If you enjoyed this article, I’d like to ask for your support. Scientific American has served as an advocate for science and industry for 180 years, and right now may be the most critical moment in that two-century history.

I’ve been a Scientific American subscriber since I was 12 years old, and it helped shape the way I look at the world. SciAm always educates and delights me, and inspires a sense of awe for our vast, beautiful universe. I hope it does that for you, too.

If you subscribe to Scientific American, you help ensure that our coverage is centered on meaningful research and discovery; that we have the resources to report on the decisions that threaten labs across the U.S.; and that we support both budding and working scientists at a time when the value of science itself too often goes unrecognized.

In return, you get essential news, captivating podcasts, brilliant infographics, can't-miss newsletters, must-watch videos, challenging games, and the science world's best writing and reporting. You can even gift someone a subscription.

There has never been a more important time for us to stand up and show why science matters. I hope you’ll support us in that mission.

Thank you,

David M. Ewalt, Editor in Chief, Scientific American

Subscribe