At some point or another, everyone has wondered how long he or she will live. Although obtaining a completely definite and reliable answer is essentially impossible, a recent study shows that your close friends may be able to provide a fairly accurate assessment of your life expectancy.
Led by Joshua Jackson, a professor at Washington University in St. Louis, a team of researchers used peer ratings and self-ratings of personality traits from the 1930s to forecast study participants’ mortality through 2013. The researchers used data from the Kelly/Connolly Longitudinal Study (KCLS), a study on personality and newly formed marriages that enrolled 600 young adults (300 heterosexual couples) between 1935 and 1938 and observed them for more than 75 years. Notably, this makes the current study the longest-running study on personality and mortality risk to date.
The KCLS participants were all recruited from the state of Connecticut and hailed primarily from middle-class backgrounds. Precise dates and places of birth were collected from all participants. At the time of the first assessment, the average age of the participants was 24.8 years. Sixty-seven percent of the participants had at least one year of college education, and 69 percent of them identified as Protestant. One limitation of the study is that all of the participants were Caucasian New Englanders. Therefore, the study sample is not fully representative of the entire North American population.
To obtain peer ratings, each participant identified three to eight friends who knew the participant well enough to provide accurate personality assessments. Five friends provided peer ratings for the majority of participants, and a total of 2,909 peer ratings were obtained. Another limitation of the study is that the study’s archival nature makes it impossible to ascertain critical characteristics of the raters other than what was first collected, such as how long and to what extent they had known the participants.
Study participants and their friends provided self-ratings and peer ratings, respectively, using the 36-item Kelly Personality Rating Scale (PRS). Previously, the researchers on the current study had validated the PRS using more modern personality measures, as the PRS is nearly 75 years old. The researchers constructed the Big Five Inventory and created equivalent factors for the PRS. Five personality traits — extraversion, agreeableness, conscientiousness, emotional stability and openness — were assessed using a five-factor solution that reflected the Big Five factor structure. This process of translating PRS ratings into Big Five ratings is comparable to the process of converting ACT scores into SAT scores.
In 1979 and 1980, data on the participants’ dates of death was collected via the participants’ children. Other primary sources used to determine dates of death were the Social Security Death Index and state death indices. As of 2013, the 21 participants who were still alive had an average age of 97.2 years. The average lifespans for men and women were 75.2 years and 81.3 years, respectively.
The results of the study demonstrate noticeable differences between men and women regarding the association of peer ratings and mortality.
For male participants, if their friends viewed them as more conscientious and open, they tended live longer. In fact, low ratings for conscientiousness were associated with the greatest risk for males of any of the traits. Agreeableness and emotional stability did not have any association with male mortality risk.
In contrast, agreeableness and emotional stability were the most important traits for female participants, while conscientiousness and openness did not factor into their life expectancy. (The extraversion trait was not correlated with mortality for either sex).
These trends persisted even after the researchers adjusted for variables such as childhood socioeconomic status, education level and household income that might otherwise affect a person’s life expectancy.
The results of the study also show differences between men and women regarding the association of self-ratings and mortality. Men’s self-ratings were significantly associated with mortality risk, while women’s were not. However, for both genders, peer ratings were much stronger predictors of mortality risk than were self-ratings.
The researchers hypothesized that the greater strength of the peer ratings as predictors of participants’ mortality was because the peer ratings were determined using responses from five different raters, whereas the self-ratings used responses from only a single rater (the participant). The researchers tested their hypothesis by examining whether the ratings from a single randomly selected peer for each participant were a better predictor of mortality than were the participant’s self-ratings. The researchers found that ratings from a single peer were no longer a significant predictor of mortality risk. This led them to conclude that the superior reliability of the peer ratings was largely due to the aggregation of ratings from multiple peers, which averaged out any idiosyncratic tendencies of individual raters.
Therefore, in order to obtain a more accurate prediction of life expectancy, rather than asking only a best friend to assess your personality, you may want to consider asking several of your close friends from several different parts of your life to do so.
