Joint Growth Trajectories of Trait Emotional Intelligence Subdomains Among L2 Language Learners: Estimating a Second-Order Factor-of-Curves Model With Emotion Perception

Abstract

The present study assessed the developmental dynamics of trait emotional intelligence (TEI) and its subdomains during English as a foreign language (EFL) learning in a longitudinal study. A sample of 309 EFL learners (217 females, 92 males) was used to assess the trajectories of the global factor of TEI and the parallel development of the TEI subdomains over 1 year in the context of the EFL classroom using parallel process modeling (PPM) and factor of curve modeling (FCM). Additionally, emotion perception (EP) was used as a distal outcome to investigate how growth parameters, including intercept and slope factors in a TEI-FCM, influence the distal outcome of EP. The results revealed that there was sufficient inter-individual variation and intra-individual trends within each subdomain and a significant increase over time across the four subdomains. Additionally, concerning the covariances within and among the subdomains of TEI, the PPM results revealed moderate to high associations between the intercept and slope growth factors within and between these subdomains. Finally, regarding the direct association of the global growth factors (intercept and slope) of TEI on EP, the results indicated that the intercept and slope of global TEI were associated with EP (γ₀ = 1.127, p < 0.001; γ₁ = 0.321, p < 0.001). Specifically, the intercepts and slopes of emotionality and sociability turned out to be significantly linked to EP (γ₀₃ = 1.311, p < 0.001; γ₁₃ = 0.684, p < 0.001; γ₀₄ = 0.497, p < 0.001; γ₁₄ = 0.127, p < 0.001). These results suggest the dynamicity of TEI during learning a foreign language are discussed in this study in light of the potential variables associated with TEI and its related literature.

Keywords: emotion perception; factor of curves modeling; longitudinal study; parallel-process modeling; trait emotional intelligence.

Figure 1

Parallel process model (PPM), a first-order longitudinal growth model (LGM). I, intercept; S, slope; TEI, trait emotional intelligence; WB, well-being; SC, self-control; EM, emotionality; SO, soociability.

Figure 2

Factor-of-curves model (FCM), a second-order LGM. I, intercept; S, slope; TEI, trait emotional intelligence; WB, well-being; SC, self-control; EM, emotionality; SO, sociability.

Figure 3

Factor-of-curves model (FCM) of TEI. I, intercept; S, slope; TEI, trait emotional intelligence; WB, well-being; SC, self-control; EM, emotionality; SO, sociability; M, mean; Var, variance; ε = error; γ = factor loading [${\chi }_{(\text{df})}^2$ = 236.84 (112), p < 0.001; CFI/TLI = 0.962/0.951; RMSEA = 0.051; SRMR = 0.055]; ***p < 0.001; **p < 0.01.

Figure 4

FCM of TEI with the a latent predictor variable (EP). I, intercept; S, slope; TEI, trait emotional intelligence; WB, well-being; SC, self-control; EM, emotionality; SO, sociability; EP, emotion perception; M, mean; Var, variance; ε = error; γ = factor loading [${\chi }_{(\text{df})}^2$ = 244.128 (98), p < 0.001; CFI/TLI = 0.974/0.963; RMSEA = 0.052; SRMR = 0.057]; ***p < 0.001; **p < 0.01.

Figure 5

A full FCM of TEI with the a latent predictor variable (EP), incorporating significant paths from EP to the subdomains. I, intercept; S, slope; TEI, trait emotional intelligence; WB, well-being; SC, self-control; EM, emotionality; SO, sociability; EP, emotion perception; M, mean; Var, variance; ε = error; γ = factor loading [${\chi }_{(\text{df})}^2$ = 261.614 (97), p < 0.001; CFI/TLI = 0.982/0.973; RMSEA = 0.053; SRMR = 0.057]; ***p < 0.001; **p < 0.01.