highly significantp < 0.001
The log likelihood ratio tests revealed that all models were highly significant compared to a single level model (Working Memory: χ 2 = 83.34, p < 0.001; Maths: χ 2 = 131.12, p < 0.001; Reading: χ 2 = 91.03, p < 0.001), indicating a substantial amount of variance attributable at the individual/upper level.