Prior findings and current research hypotheses
Although, as noted above, demographic variables have not been at the center of El studies, two of them - age and gender - have attracted sufficient research attention to warrant a brief literature review and allow our hypotheses to be based on prior findings.
Based on what is known about the development of intelligence, it has been proposed that the ability to solve emotional problems should increase with age into middle adulthood (e.g., Mayer, Salovey, Caruso, & Cherkasskiy, 2011). In the understandable absence of a long-term (life-span) longitudinal study, which could directly test this hypothesis, cross-sectional evidence of age-cohort differences has been gathered, presenting a rather complex picture. In line with the age-increase hypothesis, some studies have indeed found older participants to score higher on ability El or some of its branches (Extremera, Fernandez-Berrocal, & Salovey, 2006; Goldenberg, Matheson, & Mantler, 2006; Kafetsios, 2004; Mayer et al., 1999), yet others reported no significant effects of age on ability El (Farrelly & Austin, 2007; Roberts, Zeidner, & Matthews, 2001), and still others found these variables to be negatively related (Cabello, Navarro Bravo, Latorre, & Fernandez-Berrocal, 2014; Day & Carroll, 2004; Palmer, Gignac, Manocha, & Stough, 2005), suggesting the possibility of an age-related decline in ability El. Recently, Cabello et al. (Cabello, Sorrel, Fernandez-Pinto, Extremera, & Fernandez-Berrocal, 2016) have suggested that the apparent inconsistencies can at least partly be attributed to the methodological limitations of the above studies (restrictions in age range and overrepresentation of relatively young participants) and their sole focus on the linear effects of age on ability El. To resolve these issues, this research group recruited an adult sample with a wider age range (17-76 years) and tested not only for linear, but also for quadratic effects of age. They thereby found age-related differences in ability El across the adult lifespan to be well represented by an inverted-U curve, that is, the middle- aged group to exhibit higher ability El than both younger and older adults. Since our study focused on the adult working population - a target group with only a slightly smaller age range than in Cabello et al.’s (2016) study - we expected to find the same inverted-U relationship between age and ability EI(Hla).
Research on the relationship between age and trait El has also started off with the age-increase hypothesis, assuming that emotional self-efficacy would improve as a person matures and becomes less emotional and better socialized (Petrides, Furnham, & Mavroveli, 2007). This was indeed supported by studies which found age-related differences in trait El in favor of older age cohorts (Bar-On, 1997) or positive correlations between age and trait El (Van Rooy, Alonso, & Viswesvaran, 2005). Nevertheless, a longitudinal study, following participants from late childhood to adolescence, “revealed a complex pattern of decreases, increases, and plateaus that varied depending on age and the specific trait El domain,” but also concluded that individual differences in trait El stabilize over time (Keefer, Holden, & Parker, 2013). Supporting this conclusion, Mikolajczak et al. (Mikolajczak, Luminet, Leroy, & Roy, 2007) found no evidence of a curvilinear relationship between age and TEIQue- measured trait El (in a sample of relatively young adults), but a negligible positive correlation at the level of global scores, with only the Self-Control factor displaying a non-trivial positive association with age. In view of extant empirical findings, we expected trait El, specifically its Self-Control factor, to be positively but rather weakly related to age (H lb).