« Previous | Contents | Next »
Listen
1. INTRODUCTION
1.1 During the preliminary analysis of the Millennium Cohort Study children and families, Scotland's families and children were identified as appearing to have, on average, some distinctive characteristics from other countries of the UK. This Report takes three of these seemingly distinctive differences and provides further multivariate analysis to examine whether the distinctive results are due to differences in the characteristics of MCS families in Scotland and the rest of the UK. Scores of children's cognitive and behaviour at age 3 are analysed as well as rates of family poverty at age 3 and mothers' breastfeeding after the birth of the cohort child. All references to Tables in this Executive Summary refer to the Annex Tables.
Introduction to the Millennium Cohort Study
1.2 The second sweep of the Millennium Cohort Study ( MCS2) collected information from 1,814 families in Scotland who were part of 15,590 families of children born across the UK in 2000-2. The study's first sweep, carried out during 2001-2, when the children were aged 9-10 months old laid the foundations for this major new longitudinal research resource. It recorded the circumstances of pregnancy and birth, the all-important early months of life, and the social and economic backgrounds of the families into which the children were born.
1.3 The second survey data allow researchers for the first time to chart the changing circumstances of these children and their families and offer some direct measurements of the children's development at the age of three. Percentages reported here are re-weighted to provide representative estimates. There were differential rates of attrition by country from MCS1 to MCS2; a loss of 20% of Sweep 1 MCS families from the sample at MCS2, compared with a 15% loss from England, 17% from Wales and 22% from Northern Ireland. In addition, the loss of families from the Scotland sample was biased towards those without any educational qualifications although the rate of attrition in Scotland was higher than for England at all levels of education (Table A1.5). Fortunately MCS3 has picked up and interviewed 1444 families across the UK who were not interviewed at sweep 2.
1.4 The basic details of the sample sizes for analysis and an overview of the contents of the Millennium Cohort Study instruments are presented in Annex tables A1.1 to A1.5
The results of these analyses are all displayed in the Annex to this Report.
Weights
1.5 All of the statistics have been weighted by, in the case of Scotland, the country weight, and, in the case of the rest of the UK (England, Wales and Northern Ireland), by a specially constructed weight to reflect these 3 countries. The sample sizes given in each table are the unweighted sample sizes, unless otherwise specified. All analyses have been weighted and have had standard errors adjusted to take account of the cluster sampling design.
Topics analysed in this Report
1.6 Three topics were selected for further analysis.
(1) MCS2 children in Scotland were found to have significantly higher average scores than those in other UK countries on two cognitive assessments administered at age 3. Children in Scotland also had lower average problem behaviour scores than children in the rest of the UK. The results of the multivariate analyses of these scores are presented below ( Section 2).
(2) MCS2 families in Scotland were found to have lower rates of family income poverty at Sweep 2 than families in the rest of the UK, when children were age 3. The results of the multivariate analyses of family poverty at MCS Sweep 2 are presented below in Section 3.
(3) Lastly, MCS1 babies in Scotland were not breastfed to the same extent as babies in other countries of the UK. This information was collected when the children were aged 9-10 months old mainly from their mothers ( MCS Sweep 1). The results of analyses of whether MCS babies were breastfed are also presented below in Section 4.
Approach to the analysis
1.7 We set out to identify whether there is a genuine difference attributable to being in Scotland per se and not to identifiable characteristics of MCS families being distinctive in Scotland from other UK countries. The Millennium Cohort Study is a sample of families living in Scotland and the rest of the UK and as such it is subject to bias. It is the assessment of this bias that is being evaluated. In order to do this we carried out multivariate analyses to control for the predictors noted as significant in the respective earlier studies of each of these three topics 1 (a) children's age 3 cognitive scores, (b) family income poverty and (c) breastfeeding. MCS data offers a large and rich array of controls and predictors of child's age 3 cognitive development, family poverty and whether the child was breastfed. The earlier User Guide to the UK Millennium Cohort study had explored a range of factors separately. The intention was to explore the same factors in combination in this Report. It will then be possible to identify whether these distinctive differences at the mean for children and families in Scotland are due to differences in MCS families' characteristics which vary between Scotland and the rest of the UK and explain all the difference. Alternatively, a 'Scotland effect' may remain after controlling for these other differences in MCS families' and children's characteristics. Factors that are not in the data set, or are difficult to measure, could explain these differences, for example, they may be related to policy differences between Scotland and the rest of the UK, or cultural differences.
« Previous | Contents | Next »