Both partners’ life satisfaction and the woman’s percentage of household chores were found to be the strongest predictors of union dissolution, when researchers affiliated with Bocconi’s Dondena Center for Research on Social Dynamics and Public Policy used a machine learning (ML) technique to analyze the data. out of 2,038 married or cohabiting couples who took part in the German socio-economic panel survey. Pairs have been observed, on average, for 12 years, leading to a total of 18,613 sightings. During the observation period, 914 couples (45%) separated.
In their article, recently published online at Demography, Bruno Arpino (University of Florence), Marco Le Moglie (Catholic University, Milan) and Letizia Mencarini (Bocconi), used an ML technique called Random Survival Forests (RSF) to overcome the difficulty of managing a large number of independent variables in conventional environments. models. “A clear example of the potential difficulties of considering all the variables and their possible interactions is with the ‘big five’ personality traits,” Professor Mencarini said. “To account for the traits of both partners (10 variables) and all of their two-way interactions (25 variables), one would need to include 35 independent variables, which would be very problematic in a regression model.” ML tools are, on the contrary, able to detect complex patterns in relatively small data sets. Another advantage of ML is supposed to be its superior predictive power compared to conventional models, which are better suited to explaining the operation of certain mechanisms than to predicting the future behavior of variables. When the authors split their sample into two parts and used the results from the first half to predict the results from the second half, they found that the predictive accuracy of RSF was significantly higher than that of conventional models. Nevertheless, the predictive accuracy of the RSF was limited despite the use, as input variables, of all the most important predictors of union breakdown identified in the literature.
Among the variables with the greatest predictive power, the authors found life satisfaction of both partners, percentage of female household chores, marital status (i.e. married or cohabiting), the woman’s working hours, the woman’s level of openness and the man’s level of extroversion. The analysis also revealed that many variables interact in complex ways. For example, when male life satisfaction was high, higher female life satisfaction consistently increased the chances of union survival. But when the man’s life satisfaction was low, the association between the woman’s life satisfaction and union survival was negative after a given threshold. However, the authors did not detect an interaction effect when considering personal traits: a woman’s openness and a man’s extroversion make union dissolution more likely, regardless of regardless of their partner’s personality.
Machine Learning Predicts Marital Discord
Bruno Arpino et al, What Tears Couples Apart: A Machine Learning Analysis of Union Dissolution in Germany, Demography (2021). DOI: 10.1215/00703370-9648346
Provided by Bocconi University
Quote: Machine learning to predict if you will leave your partner (March 10, 2022) Retrieved March 10, 2022 from https://techxplore.com/news/2022-03-machine-youll-partner.html
This document is subject to copyright. Except for fair use for purposes of private study or research, no part may be reproduced without written permission. The content is provided for information only.