Language Development and Enjoyment of Reading: Impacts of Early Parent-Child Activities in Two Growing up in Scotland Cohorts

Chapter 8: Discussion and Recommendations

The Growing Up in Scotland study offers a unique opportunity to consider a range of important issues related to early years policy on supporting parent-child activities and improving children's early language development. Using data from the two birth cohorts, born six years apart, this report has considered changes in the frequency and nature of parent-child activities and in children's language ability. In particular, the report has examined whether known social differences in both of these aspects have changed over time and what may have influenced such change. In addition, data from the second birth cohort was used to measure parents' engagement with aspects of the Bookbug and PlayTalkRead initiatives - for Bookbug, receipt and use of the initial Bookbug packs; for PlayTalkRead, whether parents had accessed the website. This makes it possible to assess the penetration of these particular aspects of the initiatives amongst parents with young children in Scotland. It also allows us to see how penetration varies amongst parents with different social characteristics. Finally, it allows us to assess the extent to which engagement with these programme elements is associated with participation in home learning activities and early language development.

Children's early language development

Children's vocabulary ability increased slightly between the two cohorts. This increase remained after controlling for differences in the education levels of parents in both cohorts, and other cohort differences. The increase was seen among children from all parental educational backgrounds, with the exception of those in the lower Standard Grade group, i.e. those in the middle of the distribution. The level of increase was broadly similar in each of these sub-groups. That said there is a small but significant reduction in the vocabulary disparity between the highest and lowest social groups.

Frequency of reading at 10 months and frequency of stimulation other than reading ( e.g. painting/drawing, singing songs/nursery rhymes) at age 3 were each independently associated with children's expressive vocabulary at age 3 in both cohorts, after controlling for differences in socio-economic and other factors. This mirrors earlier findings from GUS.

The positive relationship between parent-child activities and vocabulary development is evident across both GUS cohorts. In addition, the benefits of undertaking parent-child activities, in terms of improved vocabulary ability, did not appear to differ according to parental level of education. Rather, early parent-child activities - as measured in GUS - appear to be mutually beneficial for all children.

Engagement in parent-child activities

At age 10 months and 3 years, children whose parents lived in more socio-economically advantageous circumstances were more likely than those in disadvantaged circumstances to regularly participate in home learning activities. This trend was observed in both cohorts with little change over time.

However, there were some differences in participation in individual home learning activities between the cohorts. For example, at 10 months, children in BC2 were slightly more likely than children in BC1 to be read to or to be looking at books most days. These differences remained even after controlling for other differences between the cohorts, such as parental education levels. At age 3, children in BC2 were a little more likely than those in BC1 to be playing at recognising letters, words, numbers or shapes most days. Virtually all children in both cohorts were reading or looking at books most days at age 3, with no difference between the cohorts. Taking all activities together at age 3, more than eight out of ten children, in both cohorts, had participated in frequent activities in the last week with no statistically significant difference between the cohorts.

Almost all parents who read to their child less often than 'most days' at 10 months had increased the number of days they read with their child by the time the child was aged 3. These trends were apparent and the increases similar in both cohorts. As such there is no evidence that parents in BC2 who were reading less often when the child was 10 months were more likely to have increased the frequency of reading with their child at age 3 than was the case with BC1. Of course, the proportion of parents already reading frequently at 10 months was higher in BC2 than in BC1. The remaining group - those who were not reading frequently at 10 months - being smaller in BC2, may therefore consist of a greater proportion of harder to reach parents than the equivalent group in BC1. This is a group of parents whose behaviour may be considered more difficult to influence. As such, it is perhaps unsurprising that an increase in reading frequency was not seen. This implies that more intensive efforts and support are required to influence those parents who continue to undertake less frequent reading activities with their child.

Parents in more advantaged circumstances were more likely to increase the number of days they read to their child between age 10 months and 3 years than those in disadvantaged circumstances. Furthermore, there were no notable differences between the cohorts within key sub-groups. In other words, parents with lower educational qualifications in BC2 were just as likely as parents with lower educational qualifications in BC1 to show an increase in reading frequency between 10 months and age 3.

So far we have focused on the activities themselves but we were also interested in who carried out the activities with the children. Especially noteworthy here is the increased role played by fathers across the two cohorts. There is a suggestion from the GUS data that changes in fathers' employment status may be contributing to this, at least in part. In BC2, a greater proportion of partners were working part-time than in BC1, and those who worked part-time were more likely to have engaged in all four parent-child activities than those who worked full-time. However, we have given this only rudimentary consideration in this report and further analysis is required to provide a fuller understanding of the drivers and impact of fathers' involvement in activities and why this has increased over time. With evidence increasingly suggesting the important influencing role that fathers can play in promoting positive child development, this is an encouraging finding that warrants more focused consideration.

Engagement with the Bookbug pack and PlayTalkRead website

This report has shown that parents were far more likely to be aware of and to have used the initial Bookbug pack than the PlayTalkRead website. Arguably, this is unsurprising given the difference in how the resources are accessed: Bookbug packs are predominantly distributed directly to families by health visitors and early years settings whereas the PlayTalkRead website relied largely on parents taking the initiative to access the site (and having the means to do so) in response to various marketing campaigns and promotional materials. A little over three-quarters of BC2 parents remembered receiving their initial Bookbug pack, dispatched during the child's first year. Of those who did, the vast majority used one or more of the items enclosed in the pack. By contrast, around eight in ten parents had not accessed the PlayTalkRead website at all by the time their child was aged 3.

Despite a universal approach to their rollout and availability, parents were not equally likely to have engaged with the initiatives. Parents living in the most advantaged circumstances were more likely than those in more disadvantaged circumstances to have received and used the Bookbug resources and to have accessed the PlayTalkRead website. This is not an unusual finding for resources of this nature. Parents in more advantaged circumstances, particularly those with higher educational qualifications, are known to do more research about what factors may enhance their child's developmental journey and are more likely to engage in behaviours which support positive development for that reason. As a result of this, and because more affluent parents are more likely to have the web access necessary to use the site, the social gradient is perhaps unsurprising in relation to the PlayTalkRead website; it is likely to be used more by those parents who are more motivated to seek out the resource and who have the technology to do so. Bookbug packs, however, were distributed to all parents, so explaining the social differences in receipt of these is more difficult. Of course, parents may have received their Bookbug pack many months before the point when they were asked the question and may simply not have recalled receiving it. However, if parents were regularly making use of the resources through reading or playing with their child, they are arguably also more likely to recall receiving the Bookbug pack. As such, the findings seem to suggest that parents in more disadvantaged circumstances were less likely to have received the Bookbug pack in the first place and/or were less likely to remember receiving it because they used the materials less often.

Policy initiatives and parent-child activities

A key objective of the Bookbug programme is to encourage parents to share books with their child or children from an early age. The analysis conducted for this report showed that, after controlling for other factors, parents who received or used the initial Bookbug pack read or looked at books more frequently with their child at 10 months than parents who did not receive or use the pack. When frequency of reading at 10 months was included in the model predicting frequent reading at age 3, there was no independent relationship between receipt and use of Bookbug and frequency of reading at this age. Nor was it related to an increase in reading frequency between 10 months and age 3. The association at younger rather than later ages is not surprising, given that only the initial element of the Bookbug programme was asked about. This initial element, involving a book gifted in the first few months following the child's birth, may therefore only be expected to be associated with reading behaviour in the first year. However, given that more frequent reading at 10 months was found to be associated with more frequent reading at later stages, the association between Bookbug and early reading is nevertheless notable.

It is important to emphasise that these findings do not indicate causality. That is, we cannot conclude that receipt of Bookbug resulted in parents reading more frequently with their child and that without Bookbug this would not have happened. It is possible that those parents who remembered receiving and using the Bookbug pack were also those who were already more inclined to read more frequently with their child. Yet the finding does suggest that Bookbug helps support more frequent reading behaviour before one year of age.

At the time GUS data was collected on Bookbug, the programme was universal in nature. It is perhaps unsurprising therefore, that the association between receipt and use of Bookbug and more frequent reading behaviour was equally evident amongst all parents - including both those with lower and higher levels of education. This is an important finding. There is a danger that initiatives of this nature will solely or more strongly benefit those families who are already more predisposed to engage in reading behaviour. This does not appear to be the case here.

PlayTalkRead has broadly similar aims to Bookbug in that it aims to support and encourage parents to stimulate their children from an early age through playing, talking and reading with them on a regular basis although it has quite a different approach to achieving these aims. Somewhat unexpectedly, given the limited measure of engagement with the campaign used in this report, we did find associations between accessing the PlayTalkRead website and an increase in early parent-child activities. The relationship between accessing the website and early parent-child activities was similar to that found for Bookbug: those parents who said they had accessed the website were more likely to read and sing frequently with their child at 10 months than those who had not accessed the website. The association with reading remained significant after controlling for differences in parental level of education. Again, care must be taken not to imply a causal influence here. Those parents who accessed the PlayTalkRead website may be those already more likely to engage in more frequent activities with their child. In fact, given the differences in social characteristics between those who did and did not use the site, there is some suggestion that this may be the case here.

Additional analysis suggested that the relationship between having accessed the PlayTalkRead website and the child undertaking frequent home learning activities was stronger in families where parents had higher levels of education. This bears out the risk mentioned above that initiatives of this nature, whilst intended as universal, are more likely to be used by those least in need of the advice and support they offer. The reliance on parents to be proactive in order to benefit from the PlayTalkRead website - that is, to take initiative themselves to seek out the resource and to have the means to do so ( i.e. a device to access the internet and internet access itself) - means that this particular resource is more likely to be used by those parents who are already interested in supporting their child's development, a characteristic associated with higher education levels. Analysis of engagement with more targeted aspects of the PlayTalkRead campaign may have yielded different results. However, as engagement with these targeted elements was not measured in GUS such analysis has not been possible to undertake for this report.

After controlling for other factors, children whose parents reported receiving and using the Bookbug pack had a higher vocabulary ability than those whose parents did not recall receiving the Bookbug materials. Notably, however, further analysis suggested that the relationship between Bookbug and children's language development was apparent only in cases where parents had higher levels of education. This suggests that any ultimate effect on children's vocabulary which may stem from the Bookbug programme may be less about receipt of the books per se and more about how parents use the books - and, indeed, any other books or related resources - with their child or children, reflecting different reading styles which are more or less beneficial for children's language development. As noted in the introduction, dialogic book reading - where parents are explicitly taught to actively engage their children in reading activities - has been shown to promote oral language and literacy skills. Other research has shown more highly educated parents are more likely to read up on and take advice about supporting children's development. As such, it is reasonable to assume that they may be generally more aware of the types of interaction/reading which most benefit children's language development and are more likely to engage in that style of reading.

There was no relationship between parents having accessed the PlayTalkRead website and children's vocabulary ability at age 3. With such small numbers having accessed the site (18% by the time the child was aged 3 in 2012/13), this is unsurprising. In addition, our measure of accessing the PlayTalkRead website is simply that - an indication of whether the parent was aware of and had visited the site. It does not measure the frequency of visits, the time spent considering the information, the commitment to applying the information to interactions with the child or any application as a result. Having more of this information would allow consideration of whether a certain intensity of engagement with the PlayTalkRead website is associated with better language development. Nevertheless, this would still be challenging with the small number of parents who reported using the site at all. It is also possible that a more comprehensive measure of engagement with the PlayTalkRead campaign (including the face-to-face elements) would have yielded different results.

Children's enjoyment of reading

At age eight, the vast majority of children enjoy reading and a great many like it 'a lot'. Whilst girls were more likely than boys to enjoy reading, there were no differences by socio-economic characteristics; children of all social backgrounds were similarly likely to enjoy reading. This is an encouraging finding. Whilst findings in this and other GUS reports show differences in cognitive development by social background are evident from a young age, there are no such differences in enjoyment of reading. Links between the time spent reading and positive cognitive development are well evidenced. If enjoyment of reading can therefore be translated into an increase in reading, especially among more disadvantaged children, this offers an important opportunity to address some of the inequality in cognitive development in the primary school years.

There were also no differences in enjoyment of reading according to the frequency the child was read to in their early years. Thus if a child is not frequently exposed to reading during the early years of their life this does not appear to make them less likely to enjoy reading during later childhood. However, the measure of enjoyment used here does not tell us anything about the frequency at which children read at age 8, or the extent to which they read for pleasure rather than as an exercise for school. Further research exploring these relationships would be useful in understanding more about the influence of early reading on children's later reading behaviour.

Concluding remarks

Overall, the results from this report present a mixed picture. On average, children aged 3 in Scotland in 2013 had better vocabulary ability than those aged 3 in 2007/08. Whilst there is still a large gap between the most and least advantaged children there is some suggestion that this has narrowed slightly. Most parents engage frequently in home learning activities with their children at 10 months and age 3, and the proportion of parents doing this when their child is aged 10 months has also increased over time.

These activities continue to be positively associated with better vocabulary development at age 3 for all children. This implies that if more parents can be encouraged to engage frequently in home learning activities with their child or children, this may lead to an improvement in language outcomes for children and a narrowing of inequalities between children with different social backgrounds. Furthermore, other research evidence suggests that improving children's early outcomes will impact positively on their outcomes in later life.

It is with these sorts of intentions that initiatives such as Bookbug and PlayTalkRead were developed and the evidence in this report suggests that use of each initiative was associated with a higher frequency of parent-child activities. However, the coverage or use of the initial Bookbug packs and the PlayTalkRead website amongst eligible families could be higher - particularly for the PlayTalkRead website. Our data suggests that the PlayTalkRead website favours better educated and better resourced parents. This is partly an issue of access and availability - ensuring all parents have equal opportunity to receive and use the resource - and partly about improving knowledge and awareness of the benefits - that is, ensuring the resources are acknowledged as being worth using and for the benefits they will bring. In this context it is worth highlighting, once more, that the PlayTalkRead website was only one part of a much wider campaign delivered through a combination of TV, outdoor, online advertising, social media, PR, partnerships and PlayTalkRead buses. It is possible that if wider engagement with the campaign was measured in GUS ( i.e. whether parents had seen the TV adverts, visited a PlayTalkRead bus), rather than just visiting the website, then the coverage seen amongst parents may have been higher.

Social bias is still evident in the extent to which parents engage in home learning activities with their children. It is parents in the most disadvantaged groups who are least likely to do so. These parents were also less likely to report using the Bookbug materials and the PlayTalkRead website. If the expectation is that these initiatives will help increase parent-child reading and other activities amongst the groups less likely to be doing so then it appears that a different approach should be considered for parents in more disadvantaged circumstances. The targeted Bookbug and PlayTalkRead approaches which have been launched since GUS data was collected are aimed at achieving this. Going forward, it will be important to monitor the penetration of these amongst the target groups. Even when delivered in deprived areas, there is a risk that it will be relatively advantaged families who will tend to use them, because of a greater recognition amongst these parents of the benefits of such activities. More effective communication of these benefits to all parents may improve uptake and may lead to increased home learning activities for disadvantaged children and the benefits to early vocabulary development that this will bring. Given their close involvement in the lives of many children ^[38] , it would also seem beneficial to ensure that grandparents are equally aware of the important benefits of these activities to their grandchildren.

If we return to the evidence base underpinning early book related activities discussed in the introduction to this report it is unsurprising that we do not see the sort of effects commonly seen in dialogic book reading interventions. The amount of intervention is much more tightly controlled and is specifically targeted at showing parents and especially those that are not already engaged in such activities how to get the best from book reading. Again it is not surprising given the focus of this approach that improved language skills are commonly identified as an outcome. The evidence for book gifting and the sort of generalised reading promotion that we see in Bookbug and PlayTalkRead is much more mixed because it is less the book itself than being shown how to best make use of it that is the issue. A parent driven process assumes that the parents, and especially parents who do not read to their children (10% of our population at 36 months) and may have experienced reading difficulties themselves, know what they should be doing or know where to find this information. If they do not have this information and do not know how to access it then the problem is likely to persist. Furthermore, whilst GUS data is extremely useful in providing some insight into differences in how the initial Bookbug packs and PlayTalkRead website have been used by parents, there are limits to how robustly existing data can be used to evaluate the reach and impact of national interventions such as these.

In summary, there is modest evidence of improvement in parental involvement and child outcomes over this period but it would be difficult to attribute these directly to the specific elements of the programmes explored in this report. The fact that the behaviour of parents and children who are the most disadvantaged has changed so little suggests that while universal initiatives like the Bookbug pack and the PlayTalkRead website may be helpful for engaging some parents, targeted and perhaps more creative approaches may be needed to reach the most vulnerable. We recognise that more targeted Bookbug and PlayTalkRead approaches have been launched since the GUS data was collected and that these are aimed at facilitating greater engagement from parents in more disadvantaged circumstances. For example, 'Bookbug for the Home' is targeted and delivered by professionals and volunteers who are already working with vulnerable families, and the roadshow element of PlayTalkRead, has a particular focus on areas of multiple deprivation. As noted above, it will be important to monitor the penetration of these amongst the target groups.

The analysis presented here explored parents' use of the Bookbug pack and the PlayTalkRead website among only a single cohort of parents who were exposed to the initiatives shortly after they were introduced. It has not explored whether use and reach of the resources extended as the initiatives matured.It is essential that these (and other) interventions are monitored carefully over time with particular attention to the relatively small proportion who are not engaging in book reading and other early learning activities.