MACHINE LEARNING UNCOVERS DRIVERS OF LATE-LIFE GENERATIVITY

Machine Learning Uncovers Drivers of Late-Life Generativity

Machine Learning Uncovers Drivers of Late-Life Generativity

Blog Article

With the global population aging, and rising concerns about mental health, loneliness, and social isolation, understanding and enhancing later-life satisfaction has become increasingly crucial for both individual and global health and productivity. The World Health Organization1 reports that by 2030, one out of six people will be 60 years or older, comprising 1.4 billion people, with those over 80 approaching half a billion.


Given the vast scope of this issue, it's surprising that we have a limited understanding of what preserves and enhances generativity in our later years, as research to date is still early-on. Along similar lines, increased generativity would be expected to enhance well-being, and protect against many of the negative outcomes currently associated with aging. With the global population getting older and the average human lifespan increasing, it is imperative to work out how to extend healthspan and productivity, meaning, purpose and community.

Developing Machine Learning Approaches to Understand Complex Problems


A recent study by Mohsen Joshanloo, Ph.D., published in the Journals of Gerontology (2024), took a novel approach using machine learning to extract key variables from the Midlife in the United States (MIDUS) dataset. This study included a wide range of psychological and demographic variables and measured generativity using the Loyola Generativity Scale. Participants ranged from 39 to 93 years old, with an average age of 63.64. Using maching learning allows us to make sense out of complicated data sets where standard statistical approaches may falter.

From a broad perspective, a few key concepts help us understand aging across the lifespan. These include Erik Erikson’s developmental model, especially Middle and Older Adulthood (below, with some flexibilty on the age spans given how long we're living nowadays); the balance of stability and plasticity (contributing to consistency and change); and the two major forms of well-being—eudaimonic (meaning) and hedonic (pleasure), which need to be in harmony. Personality traits, measured by the Big Five (Five Factor Model, or FFM), play a role in this process, with some contributing to stability or plasticity. For example, openness to experience is linked with plasticity, while neuroticism, because of an anxious reluctance to take risks, is often associated with stability.

The study used a "Random Forest Analysis" machine learning technique. This method builds multiple decision trees to sift through large datasets, identifying non-linear relationships and dynamic interactions. The model was trained on a subset of data and then tested against the rest to minimize prediction error and enhance accuracy. It analyzed 34 variables for 2,830 participants, after removing those with missing data and reducing the original 70 variables to a set of non-redundant measures.

The final model predicted 40% of the variance in generativity, revealing five key factors as the strongest predictors, ranked by significance:

  1. Social Potency: Assertiveness, persuasion, and leadership orientation are crucial. Generativity is inherently social, requiring not just drive but also the interpersonal skills to lead and influence others.

  2. Openness: Mental flexibility and the ability to entertain new ideas and experiences are the second strongest predictors. This is consistent with the personality trait openness to experience, which has also been linked to giftedness2.

  3. Social Integration: This factor emphasizes the importance of social function in generativity. Beyond leadership, the ability to be part of a group, experience belonging, and find one’s place in the social order is also critical. Social integration protects against loneliness, and the ill effects on health.

  4. Personal Growth: The drive for continuous development, particularly in later life, is another important contributor to generativity. As people age and face existential challenges, this drive to grow becomes even more significant.

  5. Achievement Orientation: The drive and passion to pursue goals is a key factor in generativity. A strong desire to get things done is inherently generative, though without the other factors, it may lack deeper meaning




 

Report this page