The Impact of Disability on the Lives of Young Children: Analysis of Growing Up in Scotland Data

Introduction

1.1 This research project was commissioned by Scottish Government Children and Families Analysis with the objective of undertaking an in-depth analysis of the Growing Up in Scotland (GUS) data to examine the circumstances and outcomes of children living with a disability in Scotland. The overall aim of this analysis was to explore the impact of disability on the child, their parents and the wider family unit.

Background

1.2 The Getting It Right For Every Child (GIRFEC) framework is intended to apply to all children. It is a holistic approach that requires specific needs and risks from whatever issue (whether disability, mental health, offending behaviour, medical illness, among others) to be reflected within the overarching wellbeing outcomes of SHANARRI (safe, healthy, achieving, nurtured, active, responsible, respected, included) through a child's single plan.

1.3 The GUS study offers one of the richest sources of data on Scottish children's outcomes. As yet, the data that has been collected on children living with a disability - including their early childhood outcomes - has not been subject to any in-depth analysis over and above what is systematically recorded in the annual reports and summaries on child health. The sample of children who may be classified as having a 'disability' is small, as may be expected of a nationally representative sample, which means that any data on disability cannot be disaggregated by 'disability type'. However, the sample is large enough to draw comparisons between disabled and non-disabled children.

1.4 The results of this analysis will be used to influence and inform the wider work on mapping outcomes and indicators project for disabled children and young people that is currently underway. It will also add significant value to the long term visioning of how services, support and opportunities for disabled children in Scotland can be provided and improved.

Objectives and scope of the research

1.5 The aim of this research is to use existing data to explore the characteristics, circumstances and experiences of children living with a disability in Scotland and examine how they compare with those for children without a disability. The data used will be drawn from the first six sweeps of the first birth cohort (BC1) of the GUS study. The analysis attempted to answer the following research questions:

• What are the demographics of children with a disability?
• How does the mother's experience of pregnancy and birth (with a child disabled from birth) differ from parents with a non-disabled child?
• How does disability affect the child-parent relationship?
• How is child development affected in comparison with non-disabled child developmental milestones?
• How does disability affect family structure and couple relationships?
• How does disability affect parents mental health and emotional wellbeing?
• What are parents' experiences of services for disabled children (in terms of information, usefulness, accessibility and availability)?
• What type of information/support do parents with disabled children most value?
• What are the barriers to accessing childcare and pre-school education?

Data and methods

Growing up in Scotland

1.6 The Growing Up in Scotland study (GUS) is an important longitudinal research project aimed at tracking the lives of several cohorts of Scottish children through the early years and beyond. The study is funded by the Scottish Government and carried out by ScotCen Social Research. GUS provides crucial evidence for the long-term monitoring and evaluation of policies for children, with a specific focus on the early years. The study collects a wide range of information about children and their families - the main areas covered include childcare, education, parenting, health and social inclusion. Much of the data collected is relevant to this project.

1.7 GUS launched in 2005 with two cohorts of children. The youngest of these, the birth cohort, involves a nationally representative sample of around 5217 children who were all born in 2004 or 2005. Data was collected annually from these children and their families, from the time when the cohort child was aged 10 months until they were 6 years old. Further data is being collected at age 8 and age 10.

Defining disability

1.8 The Equality Act 2010 states that "A person has a disability…if he has a physical or mental impairment which has a substantial and long-term adverse effect on his ability to carry out normal day to day activities".

1.9 The aim for this analysis was to define 'disability' based on the definition used within the Equality Act 2010, while also accounting for other definitions of disability, including additional educational support needs. A child or young person is said to have 'additional support needs' if they need more - or different - support to what is normally provided in schools or pre-schools to children of the same age.

1.10 The definition of disability for this analysis that was broadly in line with the aim outlined above and could be achieved from GUS data consistently at multiple sweeps was obtained from affirmative answers to the following question:

Does ^ChildName have any longstanding illness or disability? By longstanding I mean anything that has troubled ^him over a period of time or that is likely to affect ^him over a period of time?^[2]

And from age three onwards we have also included those who answered 'yes' to the following question also:

When we spoke to you last time you said that ^ChildName had a longstanding illness or disability. Can I just check does ^ChildName still have this longstanding illness or disability?

Analysis and presentation of results

1.11 The research questions require comparisons to be made between disabled and non-disabled children on a wide range of topics. Therefore, to answer the questions it is necessary to draw on the six existing datasets from BC1, spanning the period between birth and age 6.

1.12 Some data - such as that on the child's demographic and socio-economic circumstances - is available in all sweeps. Other data - such as that on pregnancy and birth or on parenting - is available in only a single sweep, or in a small number of sweeps. For those data which were available at each of the six sweeps, analysis is undertaken only at sweep 1 (10 months), sweep 3 (34 months/age 3) and sweep 5 (58 months/age 5). This allows a suitable consideration of differences by age whilst not presenting an unnecessary level of detail. It also allow coverage of a wider range of topics than would be the case if a single sweep was considered.

1.13 A number of research questions ask how disability 'affects' certain dimensions of family life and parenting. It is important to note that the analysis undertaken here does not demonstrate that where differences exist between disabled and non-disabled children that these differences occur as a result of the child's condition. That is, causal inferences are not possible.

1.14 It is possible however, to look at how strongly disability is associated with certain circumstances and experiences relative to, and after controlling for, other influencing factors such as socio-economic circumstances. Indeed, such analysis is crucial given the fairly strong differences in prevalence of disability according to key socio-economic characteristics as shown by Bromley^[3]. In her report, Bromley showed that children living in areas of higher deprivation and in lower income households were more likely to have had a long-term health problem in their first four years than were those living in areas of lower deprivation or in higher income households. Given that household income has been shown, in a range of other GUS research, to be related to various child and family circumstances, experiences and outcomes, it is necessary, at a minimum, to ensure that any differences between disabled and non-disabled children which are observed in the cross-tabulation analysis are not occurring due to differences in the socio-economic characteristics of children in each group.

1.15 In order to do this, multivariate analysis was used to control for differences in socio-economic characteristics and examine the independent relationship, if any, between disability and the various circumstances and outcomes of interest. Key variables of interest were converted into binary measures. For example, for parental separation, a simple measure was constructed with a value of 1 if the child's parent's had separated and 0 if they had not. These binary measures were then used as outcome variables in a series of logistic regression models. For these models, disability was added as an explanatory variable alongside a small number of standard demographic and socio-economic variables known to vary between disabled and non-disabled children and also known to influence many of the experiences of interest. These variables included: child's sex, equivalised household income (quintiles) and area deprivation (SIMD quintiles). Further information on logistic regression analysis and interpreting regression results is included in the appendix.

1.16 All figures quoted in this report have an associated margin of error, due to the fact that they are estimates based on only a sample of children, rather than all children. This margin can be estimated for each figure. For a figure which has a significance value (or p-value) of < .05 or 95%, this indicates that there is a 95% chance that the true value across all children in the population subgroup (as opposed to just those in the sample) falls within the margin. Thus a lower significance value (of < .01 or < .001) indicates a lower margin of error and a greater chance that the figure or relationship presented in the report occurs within the population. Unless otherwise stated, only statistically significant differences (between subgroups) are commented on in the text. This is true at the 95% confidence limit.

1.17 Each table provides the weighted and unweighted bases corresponding to each percentage - that is, the total number of cases on which the percentage is based. The data were weighted to compensate for differential non-response and sample drop-out across the subgroups included in GUS. Tables were created in SPSS v18 using the Complex Samples module. This module generates robust standard errors that take sample design features, such as clustering, into account. The commands identify the sample clusters; the between- and within-cluster variances are then used to generate robust standard errors.