Growing up in Scotland: changes in child cognitive ability in the pre-school years

This report examines whether the gap in cognitive ability between children from different social backgrounds changes between ages three and five and which factors influence improvement in cognitive ability.


CHAPTER 4: DIFFERENT EFFECTS BY PARENTAL LEVEL OF EDUCATION

4.1 Introduction

The analysis in the previous chapter has identified those factors, across various domains of children's lives, which impact, in both positive and negative ways, on the continuing development of their vocabulary and problem solving skills during the pre-school period. It is possible though, given the difference in circumstances and experiences of children from different educational backgrounds, that the factors which are associated with change for children with higher educated parents may differ from those associated with change for children with lower educated parents.

The final analytical stage, addressed in this section, will examine in more detail whether the effect of any of these factors varies according to parental level of education. For example, the analysis will show whether a richer home learning environment is more beneficial to children whose parents have a lower level of education, or those whose parents are more highly educated or whether there is no difference in the effect.

In this way, it is possible to identify whether the development of cognitive ability among children from the various educational groups is affected in different ways by different characteristics, circumstances and experiences and thus assess the extent to which universal or targeted policy interventions are required.

4.2 Key findings

Knowledge of vocabulary

  • Lack of educational qualifications amongst younger mothers appears to be of less importance than a lack of qualifications amongst older mothers in respect of change in relative language ability. Rather than 'doubling' the negative effect, having a younger mother in the low qualifications group appeared to cancel out some of the negative impact of the lack of qualifications.
  • The positive impact of infant-maternal attachment on improvement in relative language ability was specific to children whose parents have lower qualifications, implying that the overall negative effect on cognitive development associated with a lack of parental qualifications can be mediated somewhat by improving early infant-maternal attachment.
  • The positive relationship between early communication skills and relative improvement in vocabulary between ages 3 and 5 was more pronounced for children whose parents had no or lower qualifications. Thus children from less educated backgrounds whose relative vocabulary ability improves in the pre-school period are those who were already demonstrating more advanced communication skills at an earlier age. For children whose parents have no or lower qualifications, poor early communication skills will likely persist through the pre-school period with little or no relative improvement. In contrast, a lack of advanced early communication skills does not appear to necessarily prohibit later improvement in vocabulary for children in families where parents have higher qualifications.
  • The positive effect of attendance at ante-natal classes on relative improvement of vocabulary applied equally to children in all educational groups. Thus ante-natal classes appear to have a 'universal' positive effect. However, it is feasible that attendance at ante-natal classes is measuring characteristics associated with a desire to be a good parent which are also beneficial to the improvement of vocabulary ability in the pre-school period. Thus simply improving attendance at ante-natal classes alone is not likely to lead to improved cognitive ability.

Problem solving

  • Experiencing a higher frequency of parent-child activities such as reading, drawing, and singing nursery rhymes etc at the ages of 2 and 3 has more benefit in respect of the relative improvement of problem solving ability for children whose parents have lower qualifications than for those who had higher qualifications. Increasing such activity for children from less educated backgrounds may therefore help to reduce the ability gap.
  • None of the remaining factors associated with change in problem solving ability - attending a private pre-school, starting primary school, being breastfed and living in an area in the most deprived quintile - showed different effects for children in lower and higher educational groups.
  • Whilst it may be taken as encouraging that private pre-school attendance and early primary school experience appear to lead to a general improvement in problem solving ability, these factors do not appear to offer a policy solution for narrowing the cognitive gap and improving, in particular, the ability of those children whose parents have lower qualifications.

4.3 Education independent effects - to what degree?

In the previous chapter, Table 3.16 summarises the factors from the combined domain analysis which were associated with a change in ability in the pre-school period. Looking again at the cross-sectional analysis discussed in section 3.2, it is clear that there are stark differences in the pattern of some of the domain factors amongst the different education groups. Breastfeeding, for example, is highly socially patterned in this respect - 84% of children from households with a degree-educated parent were ever breastfed, compared with 39% of children across the two lowest educational groups. Whilst the regression analysis controls for these social patterns, it is possible that the close relationship between education and some of the explanatory variables may still be confounding the statistical process to some extent 27 .

4.3.1 Differences amongst those with higher qualifications

To explore the extent to which the various factors are associated with change in cognitive skills, beyond the effect of education, each of the final models were first run on a sub-sample of the data restricted to those parents with a degree-level qualification - this being the only education category which was shown to remain statistically significant in those final models - and then on a sub-sample restricted to those with lower level qualifications (reported in section 4.3.2). If any of the factors remain statistically significant in these degree-only models, this suggests that despite having similar level qualifications, the experience or behaviours described by the factors have an independent effect on the change in children's cognitive skills and are not simply a function of the parent's education level. The results of each model are summarised in Table 4.1 28 .

Table 4.1 Factors associated with change in cognitive ability: degree-educated parents only

Cognitive ability

Knowledge of vocabulary

Problem solving

Significant factors:

  • Language and communicative development at 22 months
  • Attendance at ante-natal classes
  • Breastfeeding

Non-significant factors:

  • Level of rule-setting in household at age 5
  • Level of infant-maternal attachment at 10 months

Significant factors:

  • Type of pre-school attended
  • Area deprivation

Non-significant factors:

  • Frequency of home-learning activities at 2-3 years
  • Whether child had started primary school
  • Breastfeeding

Of the factors which were associated with change in vocabulary for the full sample, early language development, attendance at ante-natal classes and breastfeeding remain. Thus, amongst children who each have a degree-educated parent, those who were breastfed, with better early language skills, and whose parent(s) attended ante-natal classes, were more likely to show a relative improvement in vocabulary skills in the pre-school period. For problem-solving, receiving pre-school education from a private nursery school continues to have a positive effect and area deprivation a negative effect.

4.3.2 Differences amongst those with lower qualifications

If policy is to narrow the gap in cognitive ability, then the focus needs to be on improving the performance of children from lower educational backgrounds in order that they 'catch-up' with those from higher educated households. The analysis in the previous section demonstrated that certain factors, beyond parental education, are associated with change in ability amongst children who all have a degree-level educated parent. It is possible though, given the difference in circumstances and experiences of children from different educational backgrounds, that the factors which are associated with change for children with higher educated parents may differ from those associated with change for children with lower educated parents.

To examine this, both final models were again run on a sub-sample of the full data, this time including only those cases in the lower three educational groups 29 - where parents had either no qualifications, lower-level Standard Grades (or equivalent) or upper-level Standard Grades (or equivalent). The factors which remain significant on the lower qualification group models can then be compared to those on the degree-level models allowing an appraisal of whether the effects of various factors are similar or different across the different education groups. The results of each of the lower qualification models are summarised in Table 4.2 30 .

Table 4.2 Factors associated with change in cognitive ability: lower educated parents only

Cognitive ability

Knowledge of vocabulary

Problem solving

Significant factors:

  • Level of infant-maternal attachment at 10 months

Non-significant factors:

  • Level of rule-setting in household at age 5
  • Language and communicative development at 22 months
  • Attendance at ante-natal classes
  • Breastfeeding

Significant factors:

  • Frequency of home-learning activities at 2-3 years

Non-significant factors:

  • Type of pre-school attended
  • Whether child had started primary school
  • Breastfeeding
  • Area deprivation

The results are quite different from those for the degree-level sub-sample:

  • In relation to change in vocabulary, only infant-maternal attachment remained significant; those children with higher levels of attachment were more likely to see their relative scores improve.
  • For problem-solving, the frequency of home learning activities was the only factor to return a statistically significant association. Children who experienced higher levels of these activities at ages 2 and 3 were more likely to see their problem solving skills improve in the pre-school period.

Thus, for both types of ability, amongst children from lower educated backgrounds, the importance of the home environment and parent-child relationship appears to be key. The external influences (pre-school and area deprivation) shown to have a positive relationship with change in problem solving ability in the full model, and for the degree-level sub-sample, are not evident amongst this group.

4.4 Reducing the gap - are different factors important for children with lower educated and higher educated parents?

Whilst the above analysis indicates that some factors appear to have different effects on children whose parents have different levels of education, it does not allow us to determine whether these differential effects are statistically significant. In order to do this, interaction terms between selected explanatory variables and those defining the sub-groups of parental level of education (degree or equivalent and no qualifications, lower or upper level Standard Grades) were entered into each of the main final multivariate models. Where the interaction term is statistically significant in the analysis, this indicates that the corresponding explanatory factor has a statistically significantly different effect for children in each sub-group. Table 4.3 summarises which measures were selected for inclusion in the interaction analysis and which interactions were significant. A full description of the analysis and the results is included in the Technical Appendix.

Table 4.3 Factors where an interaction with parental level of education was explored

Cognitive ability

Knowledge of vocabulary

Problem solving

Significant interactions:

  • Maternal age at child's birth: whether under 25 or not
  • Level of infant-maternal attachment at 10 months
  • Language and communicative development at 22 months
  • Whether ever breastfed or not

Non-significant interactions:

  • Attendance at all ante-natal classes

Significant interactions:

  • Frequency of home-learning activities at 2-3 years
  • Did not attend pre-school

Non-significant interactions:

  • Attended private pre-school
  • Whether child had started primary school
  • Whether ever breastfeed or not
  • Lives in an area in the most deprived quintile

In terms of circumstances and experiences which could be influenced in order to help improve the development of cognitive ability amongst children from poorer educational backgrounds, the following interaction results are notable:

  • The association between infant-maternal attachment and improvement in relative language ability was specific to children whose parents have lower qualifications implying that the overall negative effect on cognitive development associated with a lack of parental qualifications can be limited somewhat by improving early infant-maternal attachment.
  • Children from worse educational backgrounds whose relative vocabulary ability improves in the pre-school period are those who were already demonstrating more advanced communication skills at an earlier age. For children whose parents have no or lower qualifications, poor early communication skills will likely persist through the pre-school period with little or no relative improvement. In contrast, a lack of advanced early communication skills does not appear to necessarily prohibit later improvement in vocabulary for children in families where parents have higher qualifications.
  • Children whose parents have lower qualifications appear to benefit more - in terms of the improvement of their problem solving ability - from experiencing a higher frequency of parent-child activities such as reading, drawing, and singing nursery rhymes etc at the ages of 2 and 3 than those whose parents have higher qualifications.

Children vary in their cognitive ability from a very early age. At just age 3, there are stark differences in expressive vocabulary and problem solving ability according to a child's social background - household income, parental level of education and parental social class.

Longitudinal research following children from childhood into adulthood indicates that early ability is closely associated with longer-term academic outcomes and economic activity in adulthood. Research also indicates that on entering school, it becomes more difficult for under-performing children from more disadvantaged backgrounds to catch-up with their more advanced peers than is the case in the pre-school and early years period (Feinstein, 2000; Heckman et al, 2006; Sinclair, 2007). This suggests that policy efforts should be focused on maximising children's cognitive ability ahead of their school entry. The pre-school period presents a good opportunity to do so. Whilst trajectories for children with developmental vulnerabilities can be changed, the major effort is better made in the early years when the brain is most susceptible to change. Later interventions are more difficult and less effective. Maximising children's ability ahead of school entry permits them a better chance of more positive longer-term outcomes and can be done economically more efficiently than providing a 'crisis intervention' at a later stage (Scottish Government 2010b, Sinclair, 2007; Heckman et al, 2006). Thus efforts to improve cognitive ability ahead of entering school offer an opportunity for greater returns to investment and to improve outcomes over the life course.

This paper has considered two separate abilities - knowledge of vocabulary and problem solving. Whilst the existing long-term research cited above relates early ability to later educational and employment outcomes, the ability measures used are combined and development is considered more generally. Future research on the different or similar relationships that early vocabulary ability and early problem solving ability have with later outcomes would be useful in determining their relative importance in the early years.

The importance of measuring and influencing children's early development is increasingly being recognised by Scottish Government as reflected in the many recent related reports, research, policy frameworks and parliamentary debates. For example, a pilot project is currently underway to assess the usefulness of the 'Early Development Instrument' ( EDI) - a teacher-completed checklist that assesses children's readiness to learn when they enter school (Geddes and Frank, 2010). The EDI permits an assessment, at a broad population level, of how children's pre-school experiences at school entry will affect their readiness to learn at school. It can be used to influence the delivery of services and support designed to maximise children's development in the 0-5 period.

Back to top