- Top of page
- Appendix :: Mean, standard deviations, and correlation matrix
In the National Child Development Study, life-course variability in happiness over 18 years was significantly negatively associated with its mean level (happier individuals were more stable in their happiness, and it was not due to the ceiling effect), as well as childhood general intelligence and all Big Five personality factors (except for Agreeableness). In a multiple regression analysis, childhood general intelligence was the strongest predictor of life-course variability in life satisfaction, stronger than all Big Five personality factors, including Emotional stability. More intelligent individuals were significantly more stable in their happiness, and it was not entirely because: (1) they were more educated and wealthier (even though they were); (2) they were healthier (even though they were); (3) they were more stable in their marital status (even though they were); (4) they were happier (even though they were); (5) they were better able to assess their own happiness accurately (even though they were); or (6) they were better able to recall their previous responses more accurately or they were more honest in their survey responses (even though they were both). While I could exclude all of these alternative explanations, it ultimately remained unclear why more intelligent individuals were more stable in their happiness.
- Top of page
- Appendix :: Mean, standard deviations, and correlation matrix
Most empirical studies in positive psychology and the economics of happiness have analysed the level of subjective well-being, by examining who was happier than whom and what individual and social factors were associated with the level of happiness. For example, wealthier individuals were happier than poorer individuals, and people in wealthier nations were happier than people in poorer nations (Diener, Diener, & Diener, 1995), although only up to a point, beyond which additional income did not appear to produce greater happiness. Similarly, married individuals (Haring-Hidore, Stock, Okun, & Witter, 1985) and religious individuals (Ferriss, 2002) tended to have higher levels of subjective well-being.
Fewer studies have examined variability or stability in subjective well-being. Some longitudinal and panel studies tracked trends and changes in subjective well-being, either at the individual level as the life-course trends in happiness (Baird, Lucas, & Donnellan, 2010; Mroczek & Spiro, 2005) or at the societal level as the changes in aggregate mean level of happiness (Stevenson & Wolfers, 2009). Others have focused on the variability in subjective well-being over time.
In their behaviour genetic analysis of the Minnesota Twin Registry, Lykken and Tellegen (1996) estimated that happiness measured at Age 20 and at Age 30 were correlated at .50. Similarly, Schimmack and Oishi's (2005) meta-analysis showed that measures of happiness obtained 15 years apart were correlated at about .25. Fujita and Diener (2005) found in their analysis of the German Socio-Economic Panel that measures of life satisfaction over 17 years, disattenuated for measurement errors, were correlated at .34, and that the correlation between the mean and the standard deviation was significantly negative (r = −.47, p < .05, n = 3,608). Most recently, Lucas and Donnellan (2007) estimated that 34% of the variance in life satisfaction measures in the German Socio-Economic Panel Study and 38% of the variance in the British Household Panel Study were attributable to stable trait and thus did not change over time.
All of these studies on the stability of happiness, however, were univariate and descriptive. They described how stable or variable individuals' subjective well-being was over time. They did not treat the variability as an individual-difference variable; they did not explain which individuals with what characteristics were more or less stable in their happiness over time. There is one exception. Eid and Diener (1999) followed the daily fluctuations in positive and negative affect among 180 college students over seven weeks. Their analysis showed that the standard deviation in happiness had no significant bivariate correlation with any of the Big Five personality factors, but their multiple regression analysis, which included the mean, mean squared, and all Big Five factors, showed that Neuroticism significantly increased the daily variability in happiness.
At the same time, with a few exceptions (Ali et al., 2013; Isaacowitz & Smith, 2003; Siedlecki, Tucker-Drob, Oishi, & Salthouse, 2008), general intelligence has not figured prominently as a possible determinant or correlate of subjective well-being. In a comprehensive review of studies in the economics of happiness on the correlates of subjective well-being, intelligence was only briefly mentioned once as a possible unobservable trait related to education (Dolan, Peasgood, & White, 2008, pp. 99–100). The few studies that examined the influence of intelligence on happiness generally concluded that the effect was nil (Watten, Syversen, & Myhrer, 1995) or entirely mediated by demographic factors such as health and marital status (Sigelman, 1981). An international study of 192 nations showed that, of a large number of macrosocial and macroeconomic variables examined, the average level of happiness was the only factor not significantly associated with the average level of intelligence in the population (Lynn & Vanhanen, 2006). As a result, Watten et al. (1995, p. 296) concluded that ‘intelligence is virtually unrelated to QOL [quality of life]. Thus, we find IQ to be a variable of minor interest for future QOL studies.’
Siedlecki et al. (2008) showed that general (fluid) intelligence was significantly associated with life satisfaction among young and middle-aged adults, but not among older adults. This might have been because general intelligence appeared to increase both positive affect and negative affect simultaneously among the elderly population (Isaacowitz & Smith, 2003).
The purposes of the current study were twofold. First, continuing and extending the earlier work by Lykken and Tellegen (1996), Schimmack and Oishi (2005), Fujita and Diener (2005), and Lucas and Donnellan (2007), I focused on the variability and stability of happiness over time. I treated it as an individual-difference variable and explored its possible correlates. Second, I introduced general intelligence as an important factor in positive psychology in general and a significant determinant of the variability of happiness over the life course in particular.
General intelligence and stability in happiness
There are numerous theoretical and empirical reasons to expect general intelligence to be negatively associated with life-course variability in subjective well-being. Some of these factors predict that more intelligent individuals are genuinely less variable in their subjective well-being; others predict that they merely appear to be so due to some methodological or measurement reasons.
Education, wealth, and control over life circumstances
Childhood general intelligence is significantly positively associated with education and earnings; more intelligent individuals on average achieve greater education and earn more money (Brown & Reynolds, 1975; Nagoshi, Johnson, & Honbo, 1993; Snow & Yalow, 1982). Intelligence also predicts negative life events, such as accidents, injuries, and unemployment (Lynn, Hampson, & Magee, 1984; O'Toole, 1990; Smith & Kirkham, 1982). If more intelligent individuals exercise greater control over their life circumstances, because their resources protect them from unexpected external shocks in their environment, then we would expect more intelligent, more educated and wealthier individuals to experience less variability in their subjective well-being over time. Studies in positive psychology generally show that individuals return to their baseline ‘happiness set point’ after major life events, both positive and negative (Lucas, 2007). So, if less intelligent, and thus less educated and wealthy, individuals experience more negative life events, which temporarily lower their subjective well-being before they return to their baseline ‘happiness set points’, then they are expected to have greater life-course variability in happiness.
It has by now been well established in the emerging field of cognitive epidemiology that intelligence is associated with health and longevity, and that more intelligent children on average tend to live longer and healthier lives than less intelligent children, although it is not known why (Batty, Deary, & Gottfredson, 2007; Gottfredson & Deary, 2004; Kanazawa, 2006). And health is significantly associated with psychological well-being (Okun, Stock, Haring, & Witter, 1984). So, it is possible that more intelligent individuals are more stable in their happiness over time because they are more likely to remain constantly healthy than less intelligent individuals.
One of the most consistent and strongest correlates of subjective well-being is marital status; married individuals are on average happier than unmarried individuals (Haring-Hidore et al., 1985). Divorce and marriage often represent troughs and peaks of happiness in an individual's life, and intelligence is negatively associated with odds of divorce (Holley, Yabiku, & Benin, 2006). If more intelligent individuals are more likely to be consistently married throughout adulthood whereas less intelligent individuals are more likely to go through marriage, divorce, and remarriage, then general intelligence and life-course variability in subjective well-being will be negatively associated.
Mean subjective well-being
More intelligent individuals tend to be happier than less intelligent individuals (Ali et al., 2013), and the mean and the variability of life satisfaction are negatively correlated (Fujita & Diener, 2005). So, more intelligent individuals may appear to be more stable in their happiness simply because they are happier than less intelligent individuals and because the variability is an inverse function of the mean.
Greater ability to assess their own subjective well-being
Another possibility is that more intelligent individuals may be better able to assess their own level of subjective well-being more accurately at any point, so their stated level of life satisfaction remains more stable, reflecting its true stable level. In contrast, less intelligent individuals may be less able to assess it accurately, and, as a result, their verbal responses to the same survey questions tend to vary more over time at different surveys, even when their true life satisfaction remains stable.1
Another possible explanation for the effect of childhood general intelligence on life-course variability in subjective well-being involves active gene–environment interaction. More intelligent individuals may be better able to control their environment more efficiently, even above and beyond the ability afforded by their higher education and earnings, and may thus be able to live their lives more as they wish, and their lives may be less subject to unexpected environmental fluctuations, than less intelligent individuals. If this is the case, then, among other things, more intelligent individuals should be better able to predict the future states of their lives, and their future levels of happiness, than less intelligent individuals. Such greater ability to control their environment and predict its future states may also lead to more stable levels of happiness over time.
Recall accuracy and honesty
Both working memory and long-term memory are integral components of general intelligence, and more intelligent individuals on average have better memory (Unsworth, 2010). So, it is reasonable to expect that more intelligent individuals are better able to remember what their response was to the same question in an earlier survey in a longitudinal study. If more intelligent individuals are better able to recall their own responses to the same question in previous surveys, then it would allow them to appear to be more consistent, even when they are not. Lucas and Donnellan (2007) noted that this ‘autoregressive’ component accounted for about 29–34% of the variance in happiness over time.
Alternatively, it is possible that more intelligent individuals are more honest in their survey responses than less intelligent individuals. If everyone's subjective well-being is equally stable, then more honest individuals provide more consistent responses about their life satisfaction than less honest individuals, whose verbal responses fluctuate as a result of their dishonesty, even when their true level of happiness is constant.
The first three explanations above (education, wealth, and control over life circumstances; health; and marital status) propose that intelligence is a genuine correlate of life-course variability in subjective well-being, while the last three explanations (mean subjective well-being; greater ability to assess their own subjective well-being; and recall accuracy and honesty) suggest that the association between intelligence and life-course variability in subjective well-being may be due to methodological or measurement reasons.