CHAPTER 5: CONCLUSION
5.1 Differences in cognitive ability by social background characteristics
GUS data shows that the gaps in cognitive ability by social background which exist at age 3 persist at age 5. Estimates from the data suggest that, at age 5, children in more advantaged circumstances are between 11-18 months ahead in their expressive vocabulary skills and between 6-13 months ahead in their problem solving ability. The largest gaps are according to parental level of education. In terms of vocabulary, children with degree (or equivalent) educated parents are around 18 months ahead of children whose parents have no qualifications. For problem solving the equivalent gap is 13 months.
In the pre-school period, gaps in vocabulary ability by level of household income - for all children combined - decrease very slightly, and gaps by social class remain stable. In contrast, gaps between children from the lowest and highest education groups widen significantly. In relation to problem solving, the pattern is different; income gaps increase slightly, whereas gaps by education and social class decrease, quite substantially for social class. On the whole, social background factors were found to be more closely associated with, and to explain more of the variance in, vocabulary ability than problem solving ablity which may explain the different patterns of change at this broader level. Nevertheless, at this broader level, the considerable differences in cognitive ability by social background found at age 3 still remain at age 5.
Parental educational qualifications are not only a driver of the gaps in cognitive ability seen at age 5, the findings of this report suggest that parental level of education is the key social background factor which influences individual level change in cognitive ability in the pre-school period. Influences on change in ability is something that has not been much considered in other research. Children whose parents are more highly educated were significantly more likely than those whose parents had lower qualifications to see their relative performance on both assessments improve during the pre-school period. Notably, parental education was much more closely associated with the development of vocabulary skills than problem solving ability.
5.2 Other factors affecting change in cognitive ability
Many of the factors shown, through existing research, to influence early child cognitive ability - such as home learning activities and early child development - vary significantly by parental level of education. Analysis, which controlled for the effect of the level of education, showed that many of these factors continued to have an effect on change in children's cognitive ability in the pre-school years, over and above the effect of educational levels.
This analysis showed that changes in vocabulary are more related to aspects of the child's home environment and the choices and behaviours of parents. After controlling for education, greater consistency of parenting, stronger parent-child bonds, attendance at ante-natal classes and breastfeeding were each independently associated with an improvement in vocabulary ability in the pre-school years. Alongside these home experiences and parenting behaviours, early language development was also important - children who were reported to have better communication skills at 22 months were more likely to show improvement in the vocabulary ability in the pre-school period.
Change in problem solving ability appears to be more influenced by external factors such as pre-school education, although the family environment and parenting behaviours - through home learning activities and breastfeeding - were still associated with positive change. Pre-school education itself was associated with change in problem solving ability. Those children who did not attend any pre-school education were more likely to show a deterioration in problem solving ability. Interestingly, those who attended a private nursery school saw an improvement in this ability. The benefits of pre-school education to children's cognitive skills have been widely investigated and reported (Butt et al, 2007; Sylva et al, 2003). It is unclear from the analysis in this report what aspects of a private pre-school environment may cause it to be more strongly associated with a relative improvement of problem solving skills than other pre-school settings. Existing research into the effects of pre-school generally conclude that the quality of the pre-school environment - for example, measured through the warmth and responsiveness of staff towards children - is the key driver of its influence on improving outcomes. Pre-school 'practices' - such as the extent to which pre-schools particularly promote literacy or maths, or focus on positive social interactions - can also play a part. Therefore, one interpretation may be that private pre-school settings provide a better quality of environment or a different set of practices compared with other settings, and that these environments and practices are better suited to the improvement of problem solving skills than other settings. Other research has explored peer influence on individual development in the pre-school environment (Mashburn et al, 2009). The cross-sectional analysis in section 3.3.4 showed that private pre-schools were more likely to be attended by children from better educated households. Thus a further interpretation may be that the particular social mix of children in private pre-school settings contributes to its impact on problem solving ability. Further research would be useful to examine this relationship in more detail.
The proportion of children who had started attending primary school ahead of their age 5 assessment of problem solving ability appeared to benefit from that early, although brief experience. As these children do not differ in age from those who had not started school (other than in their month of birth) it is reasonable to assume this early beneficial impact will be universal - applying to all children on entry to Primary 1. Indeed, we showed no difference in effect between children from different maternal educational backgrounds. As such, this finding does not support earlier entry to school for those children at risk of poorer problem solving ability.
In the case of each ability considered, differences in prevalence of the additional explanatory factors amongst parents with different levels of qualifications explained at least some of the differences in change in ability seen by parental education rather than the qualifications per se. In other words, some of the greater change in ability shown amongst children whose parents have higher qualifications is because those children have different experiences and circumstances than children with lower educated parents. It is not simply due to parental education level itself.
5.3 Factors and associated policy responses which may help narrow the gap in cognitive ability
When the explanatory factors themselves are so closely associated with level of education, isolating their independent effect is difficult. Furthermore, it is unclear whether each factor has a similar effect on change in ability across all educational groups. 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. It is possible, given the differences 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. Thus we undertook analysis to look for any such differential effects.
We found differences for those cases where the parent was educated to degree-level and those with no or lower level qualifications showing that, in fact, the factors which matter are different for each group and in relation to each type of ability. This suggests that, at least in relation to some areas of family life and education, universal policies which seek to improve children's cognitive ability and school readiness in the pre-school period will not benefit all children equally. In fact, there are some notable findings in terms of identifying practices which may be influenced in order to improve the cognitive ability of children from lower educated backgrounds, and to reduce the cognitive gap during the pre-school period.
Having a good, early infant-maternal bond appeared to protect children in the lower education group from following the general trend of a drop-off in relative vocabulary ability amongst this group. We may assume that a good early bond will reflect, in most cases, a good continuing bond through the child's early life including in the pre-school period. This bond, and the parenting behaviours it represents, appears to compensate, at least in part, for the lack of parental qualifications and assists children from lower educational backgrounds in continuing to improve their relative vocabulary ability during the pre-school period.
Good early communication ability was also more important for children whose parents had no or lower educational qualifications. The data suggested that amongst the lower education group, those children who were reported to have better communication and language skills at 22 months were more likely to show an improvement in vocabulary over the pre-school period. This finding suggests that by assisting children from at risk groups with their communication development from the earliest possible stage would continue to have benefits to their language development throughout the pre-school years, and possibly beyond. It is imperative, however, that this intervention occurs as early as possible in order to influence emerging language skills at 22 months and is undertaken as a preventative rather than reactive measure as our findings suggest that it is particularly difficult to address existing language problems for children in this group in the pre-school period. Preventative policy is necessary because other research has indicated that disadvantaged parents are less likely to have the knowledge and skills necessary to detect early developmental delay. Furthermore, research on GUS data (Mabelis and Marryat, 2011) also shows that parents in more disadvantaged circumstances tend to have a more negative attitude to formal advice and intervention and lower formal service contact. Thus, even if some delay was suspected, parents in these circumstances may be less likely to seek help or advice. Children from these groups are also considerably less likely to attend formal childcare settings or any sort of organised parent-child group in their early years (Mabelis and Marryat, 2011). This combination of circumstances suggests that language delay amongst children from lower educated backgrounds may be less likely to be detected by a health or childcare professional than such a delay amongst children from better educated households.
For improving problem solving ability, the parent-child activities represented by the home learning environment were shown to be more important for children in the lower educational group. Experiencing - at the ages of 2 and 3 - a higher frequency of activities involving, amongst other things, reading, drawing and singing with their parents protected children in this group from the dominant trend of a decrease in problem solving ability relative to their peers. The importance of the home learning environment for the development of children's cognitive ability is not a new finding - including for analysis of GUS data (Bromley, 2009; Melhuish, 2010). However, the suggestion here is that such parent-child activities can be of particular benefit to children whose parents have lower educational qualifications. As such, current general policies aimed at encouraging parents to undertake such activities with their children - including the Play, Talk, Read campaign and the Bookbug initiative - may benefit further by providing an additional targeted element aimed at those children most likely to benefit, in terms of an improvement in problem solving ability, in the pre-school period.
As can be seen, for the most part, the factors which did suggest some variation by educational background were associated with aspects of the child's home environment and the choices and behaviours of parents - factors which are traditionally more difficult for policy to effect than more external, service-based influences such as pre-school and primary school education. However, both pre-school and early primary school experiences were shown to have some positive impact on problem solving ability. It is unclear at this point which particular characteristics of privately-provided pre-school education generate this effect. Further research is recommended to explore this so that the factors which bring about this positive change may be delivered via other pre-school settings.
The mix of family and institution effects suggests that any strategies aimed at improving school readiness via the pre-school setting will require, for more disadvantaged children, a parallel strand which seeks to influence the child's home environment and parenting experiences. This is not a surprising finding, Geddes et al (2010), in their review of interventions which are designed to improve school readiness, found that the most successful interventions utilised a mixed (centre and home-based), two generation (child and parents) approach. To ensure that children's cognitive ability is maximised in the pre-school period, our findings suggest that, in the home, such strategies should focus on the quality of the parent-child relationship and frequency of home learning activities. Delivering a support service to parents in this manner can be challenging. As Mabelis and Marryat acknowledge (2011), parents who would benefit most from such support are often those most wary of formal services, and who tend to use services less. It is important therefore, that any such support is delivered in a manner likely to engage and ensure the co-operation of those families who would benefit most from receiving it.
This report presents a complex picture of the numerous elements of children's lives which, taken together, can influence their cognitive development. Influencing just one factor is unlikely to generate any change in children's ability. Thus any policy response must recognise the multi-faceted nature of factors which impact on children's development and seek to address improvements in each of those areas in order to close the ability gap.