A new study based on the compilation of 43 datasets that consist of more than 11,000 couples suggests that perceived partner commitment is the top predictor of relationship satisfaction.
The study — published July 24 in the Proceedings of the National Academy of Science — was led by Samantha Joel, a psychology professor at Canada’s Western University; and Dr. Paul Eastwick, an associate professor at the University of California, Davis. They collaborated with seven dozen researchers.
The data for the study was collected longitudinally, meaning that researchers spoke to the same couples repeatedly at different points over time.
To determine the strongest predictors of relationship satisfaction, the researchers used “machine learning,” which the Massachusetts Institute of Technology defines as algorithms that “use statistics to find patterns in massive amounts of data.”
The study took into account several variables, including both partner’s perception of the relationship and both partners’ personality traits.
Scott Stanley, one of the collaborators on the study and research professor at the University of Denver, explained in a piece for the Institute for Family Studies that the study measured different variables such as affection, appreciation, conflict, empathy, aggression, sexual satisfaction and supportiveness.
The study also measured relationship characteristics such as whether or not couples live together, are married and how long they have been in a relationship among other characteristics.
Another category measured individual characteristics like anxiety, attachment, alcohol use, family history and demographic characteristics.
Perceived partner commitment was ranked as the top variable that explains the "variance for both present and future relationship satisfaction," according to Stanley.
According to a statement from Western University announcing the study’s conclusion, relationship-specific predictors like “perceived partner commitment,” as well as “appreciation” and “sexual satisfaction” account for nearly half of variance in relationship quality."
Perceived partner satisfaction, conflict, intimacy, love, perceived partner responsiveness, investment and trust were also among the top 10 strongest relationship-specific predictors.
Individual characteristics, on the other hand, only explained 21% of the variance in relationship quality, according to the study. The top five strongest individual predictors were life satisfaction, negative affect, depression, attachment avoidance and attachment anxiety.
“Once you have all the relationship-specific data in hand, the individual differences fade into the background,” Joel said in a statement.
According to the statement, individual differences didn’t seem to “regulate or moderate” the relationship-specific variables.
“‘Who I am’ doesn’t really matter once I know ‘who I am when I am with you,’” Eastwick said in a statement.
In an interview with CNN, Joel explained that a relationship is “more than the sum of its parts.”
"It's that relationship dynamic itself, rather than the individuals who make up the relationship, that seems to be most important for relationship quality,” Joel said.
The datasets for the study were sampled from Western countries like Canada, the United States, Israel, the Netherlands, Switzerland, and New Zealand. The researchers would like to collect data from countries in South America, Asia and Africa for future studies.
Joel and Eastwick previously published a study in 2017 titled: “Is Romantic Desire Predictable? Machine Learning Applied to Initial Romantic Attraction.” Eastwick is also the head of the Attraction and Relationships Research Laboratory, which studies “the psychological processes involved in initiating and maintaining romantic relationships.”