Current paradigms in the educational field promote education by competences with the purpose of developing in students their ability to overcome in a comprehensive way, from a know-how to do, the challenges that arise by the accelerated changes of the new century (National Education Law No. 26.206, chapter II, art. 11; Provincial Education Law No. 9870, art. 4). For De Miguel (2006), competences are a potential of behaviors adapted to a situation, an aspect entrenched in the student’s personality that can explain and/or predict their performance, and are linked to underlying characteristics such as motives, self-concept, knowledge, skills, etc. Competences will be conditioned or shaped both by the context in which their knowledge, skills, values, etc. are developed, as well as by the study situations and the students’ track of experiences inside and outside school (Moya & Luengo, 2011).

Regarding the concept of reading competence (RC), Jiménez Pérez (2014) states that it encompasses reading comprehension, which is an abstract fact that depends on the individual ability of each subject, while RC is the ability of human beings to use their reading comprehension in a useful way in the society around them. In this way, the ability is closely linked to intellectual or emotional capacities, or the individual’s psychological profile, while competence refers to a pragmatic variable, social intelligence, and socialization. On the other hand, mathematical competences (MC) entail the ability to use numbers, perform basic operations, use symbols and forms of communication of mathematical reasoning, to interpret and solve problems of daily life and work (Education Law Provincial No. 9870, art. 4). D’Amore, Díaz and Fandiño ( 2008) argue that MC are complex and dynamic. Complex, since it implies two interacting and inseparable components: use (of an exogenous nature) and domain (of an endogenous nature), in the cognitive, interpretative, and creative elaboration of mathematical knowledge in which different contents are linked and related. Dynamic, because it encompasses not only mathematical knowledge, but also metacognitive, affective, motivational, and volitional factors, being in most cases the result of diverse and interconnected knowledge. In this sense, it implies use, mastery, and availability.

Neurosciences have made significant contributions to brain functioning and the cognitive processes for learning (Dehaene, 2016; Mogollón, 2010). Likewise, other studies (Lipina & Segretín, 2015; Posner & Rothbart, 2018) have pointed out the importance of family environment, particularly the role of parents in the development of motivation to read and the acquisition of the its preconditions (i.e., development of motor skills, cognitive processes, oral language skills or abilities, and phonological awareness). Related studies also state that the competences put into play by those who fulfill parental functions have a relevant impact on the development and maturation of brain functions, as well as on the possibilities of learning, particularly school learning (Cervigni, Stelzer, Mazzoni, & Álvarez, 2012; Basuela, 2014; Fernández, 1987; Paín, 1975; Posner & Rothbart, 2018). Consistently, Mero, Pico, Rodríguez, and Bello (2015) indicated that the causes of low academic performance are very diverse, and that family variables that may be associated with other pedagogical and/or personal characteristics of the student stand out.

Research in cognitive neuroscience has also related the academic performance of students with the cognitive and affective resources placed by them in the classroom environment. Studies in this line have indicated the existence of a link between the students’ EF and their performance obtained in calculation, reading, and writing activities (Blair, 2013; Evans & Rosenbaum, 2008; Geary, Hoard, Byrd-Craven, Nugent, & Numtee, 2007). In the academic area of mathematics, studies such as those by Arán-Filippetti and Richaud (2017) have shown that WM would be involved in numerical production and mental calculation, while cognitive flexibility would be selectively associated with solving arithmetic problems. Consistently, Risso et al. (2015) and Veleiro and Thorell (2012) have shown the predictive role of WM on mathematical performance. Research conducted with elementary school students also suggests that WM and attention control play a key role in supporting emerging literacy and mathematical calculation and problem solving (Passolunghi, Vercelloni, & Schadee, 2006). Likewise, several investigations have shown the relationship between EF and RC. Among these, Arán-Filippetti and López (2016) found that WM and spontaneous cognitive flexibility (i.e., verbal fluency) would be the main executive processes that would explain, in part, individual performance variations in reading comprehension tasks. Also, Cruz and Tomasini (2005) reported the impact of self-regulation processes on reading comprehension. Previous study results have also shown that during EC tasks children create a purpose, think, review their plan, supervise it, and motivate themselves to execute it (Reyes, Barreyro, & Injoque-Ricle, 2015 ).

Overall, the competences displayed by those who fulfill parental functions have a relevant impact on the development and maturation of brain functions (Posner & Rothbart, 2018), as well as on the possibilities of learning —particularly school learning— (Barudy, 2010; Basuela, 2014; Fernández, 1987; Fernández, 2002; Paín, 1975; Vargas-Rubilar & Arán-Filippetti, 2014). However, scientific research regarding the relationship between parental competences, EF and academic competences in school-age children is scarce in the Latin American population and practically non-existent in the Argentine population. Although some research has examined the role of parenting on children’s cognitive development, there are few studies about parenting skills in processes related to learning in school-age that analyze the effect of mediating variables. Thus, this research aimed at analyzing the effect of parental skills on academic competences in school-age children, examining the mediating role of EF. Based on the aforementioned objectives, the following hypotheses were proposed:

H1. Parental competences perceived by parents have a positive and significant effect on children’s EF.

H2. Parental skills perceived by parents have a positive and significant effect on children’s academic competencies.

H3. EFs mediate the relationship between parental skills and academic competences.

We worked with a sample of 131 children of both sexes (63 girls = 48.1% and 68 boys = 51.9%), between 9 and 11 years old (M = 9.84; SD = 0.71) from public schools and their respective parents (110 women = 83.96% and 21 men = 16.04%; Mage = 36.68; SD = 7.05), citizens of Córdoba, Argentina. The criterion of inclusion was that children were not repeaters and did not have a clinical, neurological, or psychiatric history.

Perceived Parental Competence Scale (ECPP-p; Bayot & Hernández, 2008. Argentine adaptation by Azar, Vargas-Rubilar, & Arán Filippetti, 2018). The Argentine version of the ECPP-p is composed of 22 items that assess four dimensions: school involvement, dedication and orientation, shared leisure, and assumption of the role of father or mother. The instrument has four response options. Its psychometric functioning was adequate in the Argentine population (Azar et al., 2018). The analysis of the discriminative power of the items showed that the 22 items are discriminative; they obtained a score = p <.001. The reliability of the total scale from the aspect of internal consistency was highly satisfactory (α = .83). That of the dimensions ranged between low and high values (.45 to .73). The construct validity study reported a good fit with four factors: school involvement, dedication and orientation (integrating the factors personal dedication and advice and orientation of the original scale), shared leisure, and assumption of the role of father or mother (Azar et al., 2018).

Childhood Executive Functioning Inventory (CHEXI; Thorell & Nyberg, 2008). It assesses EF in children aged 4 to 12 years from the perspective of parents and/or teachers. This questionnaire has 24 items and presents five response options. It includes four subscales: working memory, planning, inhibition, and regulation. By means of EFA techniques, a two-factor solution have been found (i.e., working memory and inhibition in Swedish speaking children; Thorell & Nyberg, 2008). Both dimensions show acceptable levels of internal consistency (Cronbach’s α =.89 for WM and = .85 for inhibition) and adequate reliability (Thorell, Veleiro, Siu, & Mohammadi, 2013). A recent study also demonstrated that the CHEXI inventory is a reliable measure to assess EF in Argentinean Spanish-speaking children (Gutiérrez, Arán Filippetti, & Lemos, 2021). Regarding the parent form, the reliability measured through ω McDonald was adequate for both factors (WM ω = .86; IN ω = .84) and the total scale (ω = .92; Gutierrez et al., 2021 ).

EVALUA batteries (García Vidal, González Manjón, García Ortiz, & García Pérez, 2013 ). The EVALUA battery allows detecting school learning difficulties. It assesses general skills related to essential elements of the level of development in school age (cognitive, personal, and social), and specific skills related to dimensions of the school learning process (Reading, Writing, Mathematics, work and study habits, and academic orientation). The battery consists of two sub-batteries, EVAMAT and EVALEC. The EVAMAT 4 and 5 and EVALEC 4 and 5 batteries of the adaptation were applied (García Vidal, González, García Ortiz, & García Pérez, 2013).

EVALEC version 2.0 of the EVALUA battery (García Vidal et al., 2013). This instrument is a set of tests aimed to assess basic RC. Battery and test studies report a high level of reliability between .93 (Evalec-8) and .98 (Evalec-7). The EVALEC 4 and EVALEC 5 batteries are structured in two large blocks: reading efficiency and processes directly related to reading comprehension. The studies of the batteries and the EVALEC 4 and 5 tests report a high level of reliability, being α = 0.94 and α = 0.94 respectively (García Vidal et al., 2013). The inter-test correlation is all significant at 0.01 (bilateral). In both batteries (EVALEC-4 and EVALEC 5) and with the batteries as a whole, a high inter-test correlation of most tests stands out. This indicates a high level of homogeneity (García Vidal et al., 2013).

EVAMAT version 2.0 of the EVALUA battery (EVAMAT-4: García Vidal, García Ortiz, & González, 2013; EVAMAT-5: García Vidal et al., 2013). This instrument is designed to assess the development of mathematical skills, and it includes the following tests: (a) numbering tests; (b) calculus tests: children’s knowledge and mastery of operations and procedures to solve them; (c) geometry tests (not applied); (d) evidence of treatment of the information and chance; (e) problem-solving tests that assess skills involved in solving problem situations of a quantitative nature related to all of the above. Battery studies and EVAMAT 4 and 5 tests report a high level of battery reliability, α = .97 and α = .95, respectively (García Vidal et al., 2013). In all cases, the inter-test correlation is high, with a correlation coefficient greater than .50. In the case of EVAMAT 4, all the correlations are significant at 0.01 bilateral in both cases. A high inter-test correlation was found for most of the tests with each other, and with the batteries as a whole, which indicates a high level of homogeneity (García Vidal et al., 2013).

To collect data from parents, they were contacted by their children’s school, where they were introduced to the research objectives in a meeting organized by teachers. Then, they received the protocols, informed consents, and a note that explained the objectives and relevance of the study, as well as the ethical precautions. The evaluation of the children was carried out in groups during the usual class time.

Multivariate Analysis of Variance (MANOVA) was used to analyze differences regarding EF and performance on objective tests that assess children’s mathematical and reading competence, according to parental skills from the parents’ perspective. For said data analysis, the statistical software SPSS version 23 was used. To test different structural equation models, the AMOS version 16 was used (Arbuckle, 2007). The fit of the models was estimated using the X² test and the following Fit indices: CFI (Comparative Fit Index) and NFI (Normed Fit Index). In addition, the mean square error of approximation (RMSEA) was calculated for each model.

Actions carried out within the framework of this work considered the international ethical mandates for research with human beings (i.e., Council for International Organizations of Medical Sciences [CIOMS], World Health Organization [WHO] & Pan American Health Organization [PAHO], 2016) and the statements of the Interamerican Society of Psychology (ISP, 2019) on ethical behavior (American Psychological Association [APA], 2010; Colegio Colombiano de Psicólogos, 2009). The students participated anonymously and voluntarily, and the parents were asked to sign the informed consent to apply the instruments. No incentives of any kind were given in exchange for participation. The information collected was treated confidentially and was not accessible to anyone outside of the investigation.

Shared Leisure. An overall significant effect of this dimension on EF was found, Hotelling’s F(8, 246) = 2.635, p = .009, ηp² = .79. The Eta squared value above 0.06 indicates a medium effect. Univariate analysis indicated that there are significant differences in the four EF subscales, i.e., WM, F(2, 127) = 5.85, p = .004, ηp² = .84; planning, F(2, 127) = 4.09, p = .019, ηp² = .61; regulation, F(2, 127) = 10.01, p < .001, ηp² = .13; and inhibition, F(2, 127) = 4.86, p = .009, ηp² = .71 (Table 1). Significant differences were found between the groups with low and high levels of this PC regarding WM, planning, regulation, and inhibition. A significant difference was also found between the groups with low and medium levels of shared leisure in WM, regulation, and inhibition (Table 1).

Table 1

		Low		Medium		High		Statistics
	FE	M1	SD	M2	SD	M3	SD	F		p		η p ²
									M1 -M2	M1-M3	M2-M3
Implication School	1	21.18	6.30	19.62	5.41	20.41	7.31	0.42	0.690	0.765	0.996	0.01
	2	9.26	3.00	8.92	2.89	9.21	3.59	0.70	0.465	0.844	0.854	0.00
	3	14.47	4.10	13.74	4.39	13.74	4.42	0.14	0.861	0.996	0.922	0,00
	4	16.63	4.93	16.49	4.60	15.74	4.91	0.47	0.691	0.706	1.000	0.00
Dedication and Orientation	1	21.69	6.32	19.73	6.00	19.60	6.57	1.62	0.295	0.282	0.995	0.02
	2	9.59	2.93	8.78	3.04	8.89	3.47	0.95	0.422	0.550	0.988	0.01
	3	15.09	4.07	13.24	3.56	13.43	5.01	2.79	0.089	0.166	0.980	0.04
	4	17.39	5.35	15.66	3.96	15.57	4.65	2.18	0.089	0.166	0.997	0.03
Shared Leisure	1	22.91	6.87	19.67	5.91	18.61	5.27	5.85	0.042	0.007	0.727	0.84
	2	10.13	3.01	8.87	3.00	8.29	3.15	4.09	0.120	0.018	0.661	0.61
	3	16.15	4.29	13.20	3.63	12.58	4.02	10.01	0.002	0.000	0.780	0.13
	4	18.07	5.23	15.61	4.33	15.18	4.32	4.86	0.045	0.022	0.918	0.71
Assumption of the Father’s or mother’s role	1	20.86	6.44	22.31	5.86	18.32	6.07	3.97	0.550	0.150	0.025	0.59
	2	9.46	2.97	9.83	3.02	8.05	3.20	3.59	0.852	0.095	0.050	0.54
	3	14.63	3.92	15.46	4.16	11.92	4.12	7.95	0.638	0.007	0.001	1.11
	4	16.77	5.11	17.23	4.59	14.92	4.32	2.53	0.905	0.182	0.121	0.38

Note: 1 = Working Memory; 2 = Planning; 3 = Regulation; 4 = Inhibition.

Assumption of the Role of Father or Mother. An overall significant effect of this PC on EF was found, Hotelling’s F(8, 246) = 2.10, p = 0.036, ηp² = 0.06. Univariate analysis indicated that there are significant differences regarding WM, F(2, 127) = 3.97, p = .021, ηp² = .59; planning, F(2, 127) = 3.59, p =0.030, ηp² = .54; and regulation, F(2, 127) = 7.95, p = .001, ηp² = 1.11. No significant difference was observed regarding inhibition, F(2, 127) = 2.53, p = .083, ηp² = 0.38. Significant differences were found in the EF of WM, planning and regulation between the groups with medium and high levels of this parental competence. A significant difference in regulation was also found between the groups with low and high levels of parental competence (Table 1).

No overall significant effect for the school involvement dimension (Hotelling’s F(8, 246) = .493, p = .860, ηp² = .016) and for the dedication and orientation dimension (Hotelling’s F(8, 246) = .839, p = .569, ηp² = .027) on EF was found.

Shared Leisure. No overall significant effect was found for this dimension on RS, Hotelling’s F(12, 238) = 1.619, p = .087, ηp² = .075. However, the univariate analysis indicated that there are significant differences in terms of phonological access, F(2, 125) = 3.658, p = .029, ηp² = .055, global comprehension, F(2, 125) = 5.792, p = .004, ηp² = 0.085, and in the total of CL F(2, 125) = 3.776, p = .026, ηp² = 0.057, but not in terms of visual access, F(2, 125) = 2.522, p = 0.084, ηp² = 0.039, reading accuracy F(2, 125) = 0.075, p = .928, ηp² = .00, and literal comprehension F(2, 125) = 1.114, p = .331, ηp² = .018. Significant differences were observed between children with low and high levels of shared leisure in phonological access, between low and medium and medium and high levels in global comprehension, and between medium and high levels in the total reading competence (Table 2).

Table 2

		Low		Medium		High		Statistics
	RC	M 1	SD	M2	SD	M3	SD	F	p			η p ²
									M1 -M2	M1 -M3	M2 -M3
Implication School	1	38.25	16.43	38.57	19.96	41.01	19.66	0.26	0.996	0.768	0.842	0.00
	2	23.15	9.91	26.44	11.44	23.32	11.23	1.19	0.317	0,997	0.441	0.01
	3	16.91	6.63	16.56	6.93	18.92	4.87	1.48	0.962	0.308	0.260	0.02
	4	40.28	13.51	39.84	13.56	38.13	13.62	0.27	0.987	0.746	0.857	0.00
	5	23.17	10.37	24.05	9.60	23.72	10.90	0.08	0.914	0.968	0.990	0.00
	6	141.78	43.39	145.36	45.23	145.12	46.73	0.09	0.924	0.937	1.000	0.00
Dedication and Orientation	1	36.78	17.58	40.73	19.34	40.65	18.19	0.70	0.560	0.598	1.000	0.01
	2	22.06	11.26	25.56	10.71	25.70	9.69	1.73	0.265	0.264	0.998	0.02
	3	16.50	6.57	16.85	6.39	19.18	5.65	2.10	0.961	0.126	0.249	0.03
	4	38.55	14.64	40.10	13.88	40.55	11.25	0.27	0.850	0.777	0.988	0.00
	5	22.11	10.04	24.76	10.90	24.41	9.71	0.92	0.435	0.559	0.988	0.01
	6	136.02	45.16	147.91	48.37	150.52	37.88	1.38	0.410	0.295	0.965	0.02
Shared Leisure	1	38.45	19.34	35.36	15.35	44.22	19.39	2.52	0.697	0.318	0.071	0.03
	2	21.63	11.88	23.57	10.04	27.81	9.29	3.65	0.656	0.023	0.165	0.05
	3	17.07	6.59	17.41	6.08	17.60	6.46	0.07	0.965	0.925	0.990	0.00
	4	38.37	13.83	38.49	14.16	42.32	12.10	1.11	0.999	0.381	0.403	0.01
	5	21.94	10.85	21.36	8.76	28.12	9.87	5.79	0.957	0.014	0.006	0.08
	6	137.47	48.99	136.20	40.58	159,97	39.98	3.77	0.989	0.054	0.039	0.05
		Low		Medium		High		Statistics
	RC	M 1	SD	M2	SD	M3	SD	F	p			η p ²
									M1 -M2	M1 -M3	M2 -M3
Assumption of the Father’s or mother´s role	1	38.52	15.66	37.44	19.78	41.39	20.49	0.46	0.960	0.739	0.629	0.00
	2	24.13	10.55	21.43	11.07	26.67	10,36	2.21	0.470	0.497	0.093	0.03
	3	16.08	6.61	18.89	4.59	17.75	7.04	2.25	0.101	0.419	0.720	0.03
	4	39.71	13.15	34.90	13.93	43.71	12.42	4.08	0.214	0.323	0.014	0.06
	5	22.19	8.83	21.54	10.34	27.45	11.18	4.11	0.951	0.037	0.034	0.06
	6	140.57	39.99	134.21	46.68	156.99	46.70	2.69	0.782	0.183	0.073	0.04

Note: 1 = Visual Access; 2 = Phonological Access; 3 = Reading Accuracy; 4 = Literal Comprehension; 5 = Global Comprehension; 6 = total Reading Competence. The mean difference is significant at level 0.05.

Assumption of the Role of Father or Mother. A significant general effect of this dimension was found: Hotelling’s F(12, 238) = 1.862, p = 0.040, ηp² = 0.086. Univariate analysis indicated that there are significant differences regarding literal comprehension F(2, 125) = 4.080, p = 0.019, ηp² = 0.061 and global comprehension F(2, 125) = 4.110, p = 0.019, ηp² = 0.062. In all cases, the Eta squared value indicated a medium effect. There were no significant differences regarding visual access F(2, 125) = 0.467, p = 0.628, ηp² = 0.007, phonological access F(2, 125) = 2.211, p = 0.114, ηp² = 0.034, reading accuracy F(2, 125) = 2.254, p = 0.109, ηp² = 0.035, and in the total of RC, F(2, 125) = 2.695, p = 0.071, ηp² = 0.041. Significant differences were observed between low and high and medium and high levels of this PC regarding global comprehension, and between medium and high levels of literal comprehension (Table 2).

No overall significant effect was found for school implication, Hotelling’s F(12, 238) = 0.011, p = 0.439, ηp² = 0.049 and for dedication and orientation Hotelling’s F(12, 238) = 0.843, p = 0.606, ηp² = 0.041.

Dedication and Orientation. No overall significant effect was found: Hotelling’s F(8, 248) = 1.399, p = 0.198, ηp² = 0.043. However, univariate analyses indicated there were significant differences in numbering (F(2, 128) = 4.050, p = 0.020, ηp² = 0.060) between children with low and high levels and between children with medium and high levels of this PC (Table 3).

No overall significant effect on MC or its different dimension were found for the school involvement dimension, Hotelling’s F(8, 248) = 0.227, p = 0.986, ηp² = 0.007; the shared leisure dimension, Hotelling’s F(8, 248) = 0.893, p = 0.523, ηp² = 0.028; and the assumption of the role of father or mother dimension, Hotelling’s F(8, 248) = 1.457, p = 0.174, ηp² = 0.045.

Table 3

		Low		Medium		High		Statistics
		M 1	SD	M2	SD	M3	SD	F		p		η p 2
									M1 -M2	M1 -M3	M2 -M3
Implication School	1	27.69	9.95	27.66	9.89	29.03	10.47	0.22	1.000	0.813	0.829	0.00
	2	28.63	7.47	27.85	9.25	28.44	10.35	0.09	0.904	0.995	0.956	0.00
	3	29.49	8.78	27.52	7.71	28.77	9.74	0.60	0.518	0.922	0.813	0.00
	4	20.91	13.72	20.66	14.13	21.79	14.82	0.06	0.996	0.955	0.937	0.00
	5	106.74	29.60	103.69	29.75	108.04	34.25	0.20	0.882	0.979	0.819	0.00
		Low		Medium		High		Statistics
		M 1	SD	M2	SD	M3	SD	F		p		η p 2
									M1 -M2	M1 -M3	M2 -M3
Dedication and Orientation	1	26.92	10.17	26.11	10.15	32.02	8.65	4.05	0.916	0.045	0.026	0.06
	2	28.25	7.22	27.60	9.79	29.34	9.85	0.36	0.933	0.836	0.670	0.00
	3	28.93	8.69	27.95	8.72	29.22	8.93	0.23	0.851	0.987	0.804	0.00
	4	19.32	14.11	22.47	15.43	22.14	12.16	0.72	0.525	0.625	0.994	0.01
	5	103.44	29.61	104.16	34.07	112.73	28.08	1.10	0.993	0.344	0.447	0.01
Shared Leisure	1	26.76	10.89	27.64	9.83	30.05	9.01	1.18	0.906	0.293	0.511	0.01
	2	26.58	7.88	27.87	7.83	31.05	10.35	2.87	0.754	0.052	0.215	0.04
	3	28.59	8.38	28.24	9.60	29.41	8.10	0.19	0.980	0.903	0.813	0.00
	4	18.47	13.65	20.91	12.32	24.38	16.02	1.86	0.677	0.134	0.492	0.02
	5	100.42	28.52	104.67	29.78	114.91	33.19	2.45	0.778	0.079	0.274	0.03
Assumption of the Father’s or mother´s role	1	27.88	9.15	25.66	12.05	30.51	8.73	2.22	0.547	0.416	0.093	0.03
	2	29.17	8.35	26.19	8.68	29.13	9.35	1.49	0.249	1.000	0.322	0.02
	3	27.97	9.03	27.98	7.53	30.48	9.21	1.12	1.000	0.355	0.435	0.01
	4	19.42	11.83	20.68	15.05	23.89	15.92	1.18	0.907	0.283	0.587	0.01
	5	104.45	28.15	100.52	31.39	114.02	32.94	1.96	0.818	0.294	0.142	0.03

Note: 1 = Numbering; 2 = Calculation; 3 = Information and Chance; 4 = Problem Solving; 5 = total Mathematical Competence. The mean difference is significant at level 0.05.

Finally, different models were tested to examine the relationship between parenting skills, EFs, and academic skills. First, the effect of PCs on EFs was examined. This model showed a good fit to the empirical data, indicating that PC predicts EF from the parents’ perspective (-.28, p = .010). Model 2 (M2), which analyzes the contribution of PCs to academic competences, showed that PCs predict both Mathematical Competences (MC) (.25, p = .016) and Reading Skills (RS) (.23, p = .026). However, the model did not show a satisfactory fit to the data. In Model 3, we analyze the relationship between FE and academic competences. This model showed a good fit to the empirical data, indicating that EF predicts both MC (-.50, p <.001) and RS (-.38, p <.001). Finally, Model 4 (M4), that analyzes the direct effects of PCs on EFs and indirect effects on academic competences through EFs, showed the best fit to the data (see Table 4 and Figure 1). These results suggest that PCs have a direct effect on EF and indirect effects on academic competencies through EF.

Table 4

Model	X2	df	p	X2 /df	NFI	CFI	RMSEA
Model 1	28.508	19	.074	1.50	.94	.98	.063
Model 2	35.134	9	.000	3.90	.77	.81	.151
Model 3	15.332	8	.053	1.92	.96	.98	.085
Model 4	31.987	31	.417	1.03	.94	1.00	.016

Note: M1 = contribution of parenting skills to EF; M2 = contribution of parenting skills to reading and math skills; M3 = contribution of EF to reading and math skills; M4 = direct effects of PC on EF and indirect effects of PC on academic skills through EF.

Figure 1 Figure 1 Direct effects of parental competences on EF and indirect effects on academic competences through EF.