Publication - Research and analysis

Scottish Study of Early Learning and Childcare: ELC Leavers (Phase 2) Report - Updated 2021

Published: 27 August 2020
From: Director-General Education and Justice
Directorate: Learning Directorate, +1 more … Early Learning and Childcare Directorate
Topic: Children and families
ISBN: 9781800049567

Findings from the second phase of the Scottish Study of Early Learning and Childcare (SSELC), a research project established to evaluate the expansion of early learning and childcare (ELC) in Scotland.

Supporting documents

Child health and development

Assessments of development

Children's keyworkers at ELC settings were asked to complete observations of the child's development using the Ages and Stages (ASQ) and Strengths and Difficulties (SDQ) questionnaires. Both the ASQ and SDQ are widely used by Health Visitors across Scotland as part of their health reviews of pre-school children – the Scottish Child Health Programme^[28]. These particular instruments were also selected for inclusion in the Child Health Programme following an extensive review by academics and practitioners^[29].

The ASQ provides a structured assessment of a range of developmental domains, usually using a parental questionnaire supported by observation of the child at play, to identify children at increased risk of developmental difficulties. There are 30 items split into five different domains: communication, gross motor, fine motor, problem solving and personal-social. By answering 'yes', 'sometimes' or 'no', the respondent indicates whether or not the child can complete the action or provide the response required. Each domain produces a summary score which can be used to indicate whether the child's development is on schedule, needs monitoring or requires further assessment. Whilst it is designed to be completed by parents, because it is informed by observation of the child it was deemed suitable for completion by the child's keyworker at their ELC setting.

The SDQ is a commonly used behavioural screening questionnaire designed for use with children aged between three and 16. It consists of 25 questions about a child's behaviour to which the respondent can answer 'not true', 'somewhat true' or 'certainly true'. Responses can be combined to form five different measures of the child's development, namely emotional symptoms (e.g. excessive worrying), conduct problems (e.g. often fighting with other children), hyperactivity / inattention (for example, constantly fidgeting), peer relationship problems (e.g. not having close friends), and prosocial behaviour (e.g. being kind to others). Furthermore, the first four measures can be combined into a 'total difficulties' scale. Higher scores imply greater evidence of difficulties on each of the scales, with the exception of the prosocial behaviour scale where the reverse is true. In this report, recommended banded versions of the scales have been used to create the following categories: 'close to average', 'slightly raised', 'high' and 'very high', with 'very high' indicating multiple difficulties identified.

Ages and Stages Questionnaire

Figure 7 provides a breakdown of the proportion of children whose development was regarded as requiring further assessment, monitoring or as being on schedule across the five ASQ domains. The majority of children were considered to be on schedule on all domains, as may be expected for a nationally representative sample. The proportion of children considered to be on schedule was highest in the personal-social domain (93%), followed by the problems solving domain (89%) and the communication domain (88%), and lowest in the gross and fine motor domains (81% and 82% respectively).

Figure of child ASQ scores by domain — Figure 7: ASQ Score by domain

Base: All children (keyworker observations, weighted)

Differences in ASQ scores by gender, socio-economic status and SDQ scores

As is commonly seen on a range of different measures throughout childhood^[30], girls were more likely than boys to be on schedule on all the domains, although the large majority of boys as well as girls were on schedule in each case. On three of the domains (communication, fine motor and problem solving), there were associations between area deprivation and not being on schedule, and between low income and not being on schedule. On the gross motor, fine motor and problem solving domains there were also associations between low levels of education and not being on schedule. Scores on the personal-social domain were not associated with any of these measures of socio-economic status. Further details of these associations are provided in the following sections.

It is worth noting that socio-economic factors tend to be associated with each other: those with low levels of education are more likely to be on lower incomes and to live in more deprived areas. Certain risk or protective factors also tend to be more prevalent in different socio-economic groups. For example, breastfeeding and high frequencies of home learning activities are both associated with high levels of parental education. The differences observed in the following sections looking at socio-economic factors should therefore be viewed in light of the other figures discussed later looking at the child's home environment. Any difference seen between sub-groups is not necessarily due to the particular factor being considered, but could be due to a combination of factors. In addition, in some cases there may be no causal association at all between a specific risk factor and the ASQ scores, with the association due entirely to other known or unknown factors. Towards the end of the section on child outcomes we will look at regression analysis to determine the main drivers behind the observed associations.

There were also very strong correlations between ASQ and SDQ scores. While both measure child development, they do measure different things. The SDQ is concentrated on social, emotional and behavioural development, while the ASQ is more broad in its coverage of developmental milestones. There are many children picked up by one of the measures as potentially exhibiting development that is behind schedule, but not assessed as at risk on other measures. At the same time there are many children whose development is behind schedule in multiple different ways.

ASQ communication domain

Figure 8 breaks down scores on the ASQ communication domain by the sex of the child, the child's SDQ score and by area deprivation (SIMD). Girls were more likely to be on schedule (90%) than boys (86%), although the difference for the communication domain is not as large for some of the other domains. Children living in the most deprived areas were less likely to be on schedule (85%) than children living in other areas (89%). Children from wealthier families were also more likely to be assessed as on schedule than those from less wealthy households (94% of those in the top two income groups, compared with 84-86% of those in lower income groups) (see Appendix Table C6). No significant differences were noted by education or by number of parents.

As was the case with the other ASQ domains, a child's score on the communication domain was correlated with their SDQ total difficulties score. Those who had a SDQ score that was close to average were much more likely to be on schedule in the ASQ communication domain (92%) than those who scored 'very high' on the SDQ total difficulties scale (57%).

Figure of child ASQ communication domain scores by sex, SDQ total difficulties score and SIMD — Figure 8: ASQ communication domain by child's sex, SDQ total difficulties score and SIMD

Base: All children (keyworker observations, weighted)

ASQ gross motor domain

Figure 9 provides a breakdown of the assessments in the gross motor domain by sex of the child and their SDQ total difficulties score. As was the case across all ASQ domains, boys were more likely to be assessed as requiring further assessment or monitoring than girls. Development was assessed as being on schedule for 87% of girls compared with 76% of boys. Those who scored close to average on the SDQ total difficulties score were also more likely to be on schedule in the ASQ gross motor domain than those who scored 'very high' (86% compared with 46%). There were no significant differences in gross motor scores by area deprivation or by number of parents, although there were differences by highest level of parental education. Where the respondent (generally the mother) had at least Highers or equivalent upper school or post-school qualifications, the child was more likely to be on schedule (82-85%) than where the respondent only had lower school qualifications (such as Standard Grades or equivalent) or no qualifications (72-74%). (See Table C7 in Appendix C).

Base: All children (keyworker observations, weighted)

ASQ fine motor domain

Some of the largest differences between subgroups were evident in the ASQ fine motor domain. The gap between boys and girls as shown in Figure 10 was highest in this domain, with 92% of girls being assessed as on schedule for fine motor skills, compared with 73% of boys. Children living in the most deprived areas were less likely to be on schedule (76%) than those living elsewhere (85%), and children who were close to average on the SDQ total difficulties score were much more likely (86%) to be on schedule than those who scored 'very high' (41%).

The child's score on the ASQ fine motor domain was also associated with all of the other socio-economic characteristics examined, as shown in Tables C8 to C10 in Appendix C. Higher household income and higher levels of parental educational attainment were both associated with increased likelihood of a child being assessed as on schedule for fine motor development, as was living in a two parent household (compared with a single parent household).

Figure of child ASQ fine motor domain scores by child’s sex, SDQ total difficulties score and SIMD — Figure 10: ASQ fine motor domain by child's sex, SIMD and SDQ total difficulties Score

Base: All children (keyworker observations, weighted)

ASQ problem solving domain

For the ASQ problem solving domain a number of associations with socio-demographic factors were again evident. Girls were more likely than boys to be on schedule (93% compared with 85%). Those in the most deprived areas (84%) were less likely to be on schedule than those residing in other areas (91%). Figure 11 also illustrates the strong association between the ASQ problem solving domain and the SDQ total difficulties score, with 92% of those with a close to average SDQ total difficulties score assessed as on schedule for problem solving, compared with 64% of those who scored 'very high'. The appendix tables reveal further associations between household and socio-economic factors and increased likelihood of being on schedule for problem solving, including living in a high-income household (Table C11), having parents with at least Highers or equivalent (Table C12) and living in a two-parent household (Table C13).

Base: All children (keyworker observations, weighted)

ASQ personal-social domain

Finally, Figure 12 breaks down the ASQ personal-social domain by sex and SDQ total difficulties score. Girls (96%) were more likely to be on schedule than boys (90%) and those with close to average SDQ scores (97%) were more likely than those with 'very high' scores (59%) to be on schedule. Socio-economic status did not appear to be associated with scores on the ASQ personal-social domain.

Base: All children (keyworker observations, weighted)

Strengths and Difficulties Questionnaire

Figure 13 provides a breakdown for all five of the SDQ domains as well as the total difficulties score. The prosocial behaviour domain is coloured differently to highlight the different labelling, although the general meaning of the categories are similar. The scales are designed so that most children have a score in the 'close to average' range, with around 20% of the population having raised scores (lowered for prosocial behaviour), although some variation is expected according to the age of the children. Data from the keyworker observations shows that for each of the domains, proportions of children with raised scores are as expected from a nationally representative sample with 85% of children scoring close to average in their SDQ total difficulties score (the sum of the scores for the four "problem" domains). The domain in which the most children scored close to average was prosocial behaviour (92%), followed by conduct problems (89%) and peer problems (86%). The domains in which the lowest proportion of children scored close to average were hyperactivity (78%) and emotional symptoms (80%).

Figure of child SDQ scores by domain — Figure 13: SDQ scores by domain

Base: All children (keyworker observations, weighted)

Differences in SDQ scores by child's sex and SIMD

As seen with the ASQ scores, the child's sex was a significant factor in predicting child outcomes. Across all the SDQ domains, with the exception of emotional symptoms, girls tended to have fewer difficulties (indicated by raised scores on the problem domains and lowered scores on the prosocial behaviour domain) than boys (see Figure 14). For the total difficulties score, 91% of girls were close to average compared with 79% of boys, with 7% of boys scoring 'very high' compared with 2% of girls. The gap was largest for the hyperactivity domain (88% of girls close to average compared with 69% of boys). The gap was smaller, but still significant, for the conduct problems domain (84% of boys and 93% of girls close to average), the peer problems domain (83% of boys and 88% of girls close to average), and the prosocial behaviour domain (89% of boys and 96% of girls close to average). Differences between boys and girls were not significant for the emotional symptoms domain.

Area deprivation was not significantly associated with a child's SDQ total difficulties score, nor was it for any of the individual domains.

Figure of child SDQ domain scores by child’s sex — Figure 14: SDQ domain scores by child's sex

Base: All children (keyworker observations, weighted)

Differences in SDQ total difficulties score by socio-economic status

To avoid further repetition, the remaining analysis of the SDQ scales has been conducted for the total difficulties score only. As shown in Figure 15 and Tables C14 and C15 in Appendix C, a child's SDQ total difficulties score was also associated with a range of socio-economic factors. Children were more likely to have raised scores if they lived in low-income households (Figure 15). Around one-in-five (20-21%) of those in the two lowest income quintiles had raised / high scores, compared with one-in-ten (7-11%) of those living in higher income households. Similarly, children were more likely to have raised scores if their parent had no formal educational qualifications (Table C14) or if they lived in a single parent household (Table C15).

Figure of child SDQ total difficulties scores by OECD equivalised household
income — Figure 15: SDQ total difficulties score by OECD equivalised household income

Base: All children (keyworker observations, weighted)

General health and long-term illnesses

Parents were asked to rank their child's general health as either 'very good', 'good', 'fair', 'bad' or 'very bad'. Three-quarters of parents (76%) ranked their child's health as being 'very good', and a further one-fifth (21%) ranked it as being 'good'. Only 2% of parents stated their child's health was 'fair' and fewer than 1% ranked it as being 'bad'.

There was no noticeable difference in perceptions of the child's general health for boys and girls. Nor was there a significant difference in perceptions of children's general health related to area deprivation. However, there was a difference in perceptions between single parent and couple parent households with 65% of those in single parent households perceiving their child's health as being 'very good' compared with 78% of those in couple households (see Table 5).

Table 5: Child's general health by household type

	Single parent	Couple parent	All
	%	%	%
Very good	65	78	76
Good	29	20	21
Fair	6	2	2
Bad	-	0	0
Unweighted base	274	1,096	1,380

Base: All respondents (parent survey, weighted)

One-in-ten children (10%) had a longstanding illness or health condition. Of these children, 15% of parents said that the longstanding illness limited the child 'a lot', 44% said it limited them 'a little' and 41% stated the longstanding illness did not limit their child, meaning that 6% of all four- and five-year-old children attending an ELC setting had a longstanding limiting illness. Boys were more likely to have a longstanding condition than girls (12% compared with 8%). There was also a difference by household type with 14% of children in single parent households having a longstanding illness compared with 9% of children in two parent households.

Longstanding illnesses were most commonly associated with mental health, social, emotional or behavioural issues (30% of those with a long-term condition). However, one quarter of those whose child had a longstanding illness stated it affected the child's stamina or breathing and just over one-fifth (22%) said it affected the child's learning, concentrating or remembering. Table 6 provides a complete summary of these responses. Seventy-one percent of the parents of children with a long-term condition mentioned that it affected their child in just one of the ways listed in Table 6, while 21% mentioned two different ways and 8% three or more.

Table 6: How longstanding condition or illness affects child*
	All
	%
Vision	13
Hearing	11
Mobility	10
Learning, concentrating or remembering	22
Stamina or breathing difficulty	25
Mental health, social, emotional or behavioural issues	30
Other impairment(s)	30
Unweighted base	128

Base: All children with a long-term condition (parent survey, weighted)

*Note: respondents were able to choose more than one response. As such, percentages will not total 100%.

Parents were asked whether they had any concerns about how their child talks or how they understand. The majority of parents had no concerns about either (86% and 96% respectively). There were some differences between boys and girls, as shown in Figure 16. For girls, 99% of parents had no concerns about how they understood, compared with 93% of parents of boys. Seventeen percent of parents of boys had concerns about how their child talks compared with 11% of parents of girls. There were no significant differences in responses between single and couple parent households, nor was area deprivation a significant factor.

Figure of parental concerns about how the child talks and understands by
child’s sex — Figure 16: Parental concerns about how the child talks and understands by child's sex

Base: All respondents (parent survey, weighted)

Home environment

Parents were asked for a range of information capturing aspects of their child's early childhood circumstances, experiences and behaviours which are known to be associated with child development. These include sleep patterns, breastfeeding, and activities at home that promote learning. Once we have explored these in more detail, we shall return to the assessments of development to examine the relationships between ASQ and SDQ scores and the child's home environment.

Six-in-ten children aged four or five (60%) slept through the night every night (Table 7). There were no significant differences in this measure according to sex of the child, area deprivation or household type.

Table 7: How often child sleeps through the night without waking
	All
	%
Never sleeps right through the night	6
1-2 times a week	7
3-5 times a week	13
6 times a week	14
Every night	60
Unweighted base	1,374

Base: All respondents (parent survey, weighted)

Just over half (55%) of children had ever been breastfed, with no significant differences between boys and girls. There were differences in incidences of breastfeeding related to area deprivation and household type, summarised in Figure 17. Only one-third (34%) of children who lived in the most deprived areas had ever been breastfed, compared to more than two-thirds (70-71%) in the two least deprived quintiles. There was also a lower rate of breastfeeding among single parents than couple parents: 37% of single parents reported that their child was ever breastfed, compared with 59% of those living in two-parent households.

Figure of whether child was ever breastfed, by SIMD and household type — Figure 17: Whether child was ever breastfed by area deprivation ( SIMD) and household type

Base: All respondents (parent survey, weighted)

Parents were asked how often the child had been engaged in various home learning activities in the previous seven days. These included looking at books / reading stories, painting or drawing, reciting nursery rhymes / singing songs and playing at recognising letters, words, numbers or shapes.

The most common activity was looking at books or reading stories which 57% of children had done every day in the previous week. Reciting nursery rhymes / singing songs and recognising letters, words, numbers and shapes were also relatively common activities with 51% of children having done the former every day in the last seven and 42% the latter. Painting and drawing was less common, with a quarter (25%) of children having done this every day. Thirteen percent of children had done all four of the activities on every day in the previous week, while very few (< 0.5%) had done none of the activities at all over that time.

There were some differences in activities engaged in between girls and boys, between those in single and couple parent households, and according to area deprivation, as shown in Table 8. Boys were less likely to have been engaged in activities related to painting and drawing than girls - 16% of boys having done this each day in the last seven compared with 34% of girls. Boys also sang songs / recited nursery rhymes less frequently (45% of boys did this every day compared with 58% of girls). Children in single parent households were less likely to have looked at books and read stories than children in couple parent households. Thirty-nine percent of those in single parent households had read books / looked at stories every day in the last week compared with 61% of children in couple parent households. Those living in deprived areas were also less likely to have looked at books every day: 41% of children in the most deprived areas had done so, compared with 61% of children in other areas.

Table 8: Frequency of home learning activities by child sex, area deprivation and household type
	Male child	Female child	Most deprived 20%	Other area	Single parent	Couple household	All
	%	%	%	%	%	%	%
Looked at books or read stories
Not in last 7 days	2	1	5	1	3	1	2
1 to 3 days	16	14	23	13	29	12	15
4 to 6 days	27	26	31	25	30	25	26
7 days	55	58	41	61	39	61	57
Painting or drawing
Not in last 7 days	6	1	4	3	3	3	3
1 to 3 days	45	23	34	35	36	34	34
4 to 6 days	33	42	40	37	38	37	37
7 days	16	34	22	25	22	26	25
Recited nursery rhymes or sung songs
Not in last 7 days	2	1	2	2	3	1	2
1 to 3 days	21	13	17	17	17	17	17
4 to 6 days	31	28	25	31	30	30	30
7 days	45	58	57	50	50	52	51
Recognising letters, words, numbers or shapes
Not in last 7 days	2	2	3	2	5	1	2
1 to 3 days	20	19	22	19	22	19	20
4 to 6 days	35	38	37	36	33	38	36
7 days	43	41	38	43	40	42	42
Unweighted base*	684	673	408	236	267	1,086	1,361

Base: All respondents (parent survey, weighted)

*Note: bases vary slightly for each individual activity. The bases shown are the lowest amongst the four activities.

The frequencies at which children had done each activity in the last seven days were summed together to create a total home learning activity score. The score ranged from 0 to 28 with the highest score indicating the child had been engaged in all four activities on each of the previous seven days. The cohort was then split into four evenly sized groups (quartiles) according to their scores on the scale: 0 to 16 (24% of all children); 17 to 21 (28%); 22 to 24 (23%) and 25 to 28 (26%).

Figure 18 breaks down the home learning environment (HLE) score quartiles by the parent's highest educational qualification.

Figure of home learning environment scores by parent/carer education level — Figure 18: Home learning environment score by parent's highest educational qualification

Base: All respondents (parent survey, weighted)

In general, the higher the educational qualification held by the parent, the more likely the child was to be in a higher HLE group. Half of those whose parents had no qualification were in the lowest quartile for HLE, representing the least frequent home learning activities, compared with 18% of those whose parents had a degree or equivalent. Only 12% of those children whose parents had no qualifications were in the top quartile for HLE, compared with 30% of those with post-school qualifications (HNC, HND, degree or equivalent).

Parents were also asked about how chaotic or ordered their home was using four items taken from the Confusion, Hubbub and Order Scale (CHAOS)^[31]. The items ask parents how much they agree or disagree with four statements:

It's really disorganised in our home
You can't really hear yourself think in our home
The atmosphere in our home is calm
First thing in the day, we have a regular routine at home

Responses to these questions are summarised in Figure 19.

Figure of extent of parent/carer agreement with household order / CHAOS items — Figure 19: Agreement and disagreement with household order / chaos items

Base: All respondents (parent survey, weighted)

Most parents disagreed with the first two statements and agreed with the last two, indicating that their homes were calm and organised. Overall, 91% agreed that they had a regular routine first thing in the day, 70% agreed that the atmosphere in their home was calm, 74% disagreed that they could not really hear themselves think in their home and 83% disagreed that it was really disorganised in their home. There was no significant variation in responses to any of these statements by sex of the child, area deprivation or by household type.

Responses to these four statements were summed together - reversing the last two items - to create a scale with possible values from 4 to 20, with higher values representing more chaotic households. The scale was then divided into three groups of equal size indicating households with low, moderate and high levels of chaos. This scale was used earlier as a predictor of ASQ and SDQ scores. There was no significant variation in this scale according to sex, area deprivation, household type, parental education or income.

Assessments of development and home environment

The ASQ domains and the SDQ total difficulties score were examined in relation to a number of risk factors or family attributes (see the sections on developmental risk factors, parental outcomes and characteristics of the cohort for more details of these measures). In this section we look first at the associations between individual risk factors and child development outcomes, and then use logistic regression analysis to identify the key drivers of development outcomes.