highly significantp < .001
In accordance with previous studies, there was a highly significant inverse bivariate association of parental SEP with AL ( r = −.26, p < .001) (Additional file 3 : Table S2: Correlation matrix of all study variables).
In accordance with previous studies, there was a highly significant inverse bivariate association of parental SEP with AL ( r = −.26, p < .001) (Additional file 3 : Table S2: Correlation matrix of all study variables).
While the direct effect of social relations satisfaction on AL was marginally significant ( β = − 0.094, p = .066), neither social relations satisfaction nor dissatisfaction significantly mediated the parental SEP-AL association.
In the present study, using the WAIS Full Scale IQ score allowed us to test the mediating role of such a construct; while there was a borderline significant indirect effect of intelligence in a model with this as the only mediator, this effect was strongly attenuated in Model 3 when including years of education.
This strengthened the association of social relations satisfaction with AL in Model 2 ( β = − 0.103, p = .046) and slightly attenuated the direct association of parental SEP with AL in all models; in Model 3 ( β = − 0.100, p = .068) and model 4 ( β = − 0.097, p = .064) to marginal significance.