03
Oct
2022

Feelings integers are highly predictive of future human behaviour, research shows

measurement cause and effect

New research shows that a person’s own rating of their feelings – even on a seemingly arbitrary scale – is of greater predictive power than a collection of socioeconomic measures.

The findings, published today (Monday) in Proceedings of the National Academy of Sciences, were made by researchers at the Wellbeing Research Centre, University of Oxford, using data from approximately 700,000 people across multiple countries.

Professor Andrew Oswald (Warwick) and Dr Caspar Kaiser (Oxford) examined the relationship by comparing self-reported feelings integers – for example, where individuals were asked to rate their satisfaction on a scale of 0 to 10 – to later ‘get-me-out-of-here’ actions.

These actions, where individuals choose to leave their current setting, are an unambiguous signal of human dissatisfaction with the status quo. For the purposes of this study, the authors looked at four types of get-me-out-of-here action: moving dwellings, changing intimate partners, leaving jobs, and hospital visits.

Across 34 years of data in Germany, 25 years in the UK and 20 years in Australia, their research shows that feelings integers are generally of greater predictive power than combined socioeconomic variables including household income, marital status, education and number of children, among others.

The researchers describe a stable and almost linear relationship between a single feelings integer and these self-driven life changes, in all three of the countries examined in the study.

Dr Caspar Kaiser, Research Fellow at the Wellbeing Research Centre and a Research Officer at Oxford’s Institute for New Economic Thinking (INET), and corresponding author for the study, added: “It is unknown whether our results will replicate more globally, especially in low- and middle-income countries. Another interesting next step would be to examine whether the observed action-satisfaction associations systematically differ across population groups, e.g. between men and women or across age.”

The scientific value of numerical measures of human feelings’ is published in Proceedings of the National Academy of Sciences of the United States of America.