Scottish Longitudinal Study
Development & Support Unit
TECHNICAL WORKING PAPERS
Guide to parallel and combined analysis of the ONS LS, SLS and NILS
Young, H. (2009) CeLSIUS: London. 1 July 2009. [SLS][ONS LS][NILS]
The Office for National Statistics Longitudinal Study (ONS LS), covering England and Wales was established in the mid 1970s and now contains data from up to four Censuses, together with information on vital events currently (July 2009) available from the 1971 Census until the end of 2006. In 2007 the Scottish Longitudinal Study (SLS) and the Northern Ireland Longitudinal Study (NILS) were both launched; both, like the ONS LS, include information from Census and vital registration sources, although there are a number of differences in linked data available which are referred to below. This means that it is now possible to carry out UK-wide analysis using Census based record linkage studies.
The aim of this document is to provide a guide for those who wish to carry out analysis of two or three of these datasets together. The information presented is based on experience of carrying out analysis on the three studies together as part of an ESRC Census Programme funded project. The next section briefly presents similarities and differences between the three datasets. Subsequent sections document legal and logistical issues in carrying out combined analyses; options for project design; procedures for the application process; and issues in data preparation and analysis.
Download output document: Technical Working Paper (PDF 59KB)
THE SCOTTISH LONGITUDINAL STUDY, An Introduction
Hattersley, L. & Boyle, P. (2007) SLS Technical Working Paper 1. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 16 April 2007. [SLS]
The Longitudinal Studies Centre – Scotland (LSCS) is an ESRC-funded centre of excellence for the creation, manipulation and analysis of longitudinal data. Its main project has been the establishment of the Scottish Longitudinal Study (SLS) which is a large scale linkage study created from the linkage of data from routine administrative and statistical sources. These include Census data, vital events data (births, deaths, marriages), National Health Service Central Register (NHSCR) data (migration in or out of Scotland) and NHS data (cancer registrations and hospital discharges). ...
Download output document: SLS Technical Working Paper 1 (PDF 1MB)
THE SCOTTISH LONGITUDINAL STUDY, Tracing rates and sample quality for the 1991 Census SLS sample
Hattersley, L., Raab, G. & Boyle, P. (2007) SLS Technical Working Paper 2. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 1 October 2007. [SLS]
This working paper covers the creation, selection and quality of the 1991 Census SLS sample that was used as the starting point of this study. It provides details of the methods used to select the sample from the 1991 Census returns and the process of flagging the study members on the NHSCR system. It goes on to discuss the quality of the tracing rates and sampling fractions. This is designed to be a technical report and for a more general introduction to the SLS, please refer to LSCS Working Paper 1 “The Scottish Longitudinal Study: an introduction”.
Download output document: SLS Technical Working Paper 2 (PDF 337kb)
THE SCOTTISH LONGITUDINAL STUDY, A technical guide to the creation, quality and linkage of the 2001 Census SLS sample
Hattersley, L. & Boyle, P. (2008) SLS Technical Working Paper 3. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 1 November 2008. [SLS]
...This working paper covers the creation, selection and quality of the 2001 Census SLS sample. It provides details of the methods used to select the sample from the 2001 Census returns and also discusses the effect on the SLS sample of the particular methodology used in the 2001 Census (known as the ‘One Number Census’ methodology). ...
Download output document: SLS Technical Working Paper 3 (PDF 328KB)
THE SCOTTISH LONGITUDINAL STUDY, The 1991 – 2001 Scottish Longitudinal Study Census Link
Hattersley, L. & Boyle, P. (2009) SLS Technical Working Paper 4. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 1 April 2009. [SLS]
...This working paper covers the linkage of the 1991 and 2001 SLS Census samples. It examines the quality of the linkage and explains why linkages may not be possible in some cases. ...
Download output document: SLS Technical Working Paper 4 (PDF 258KB)
Geographies and ecological variables in the Scottish Longitudinal Study (SLS)
Feng, Z. (2013) SLS Technical Working Paper 5. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews. 28 May 2013. [SLS]
The Scottish Longitudinal Study (SLS) is a large scale linkage study created from the linkage of data from routine administrative and statistical sources. These include Census data, vital events data (births, deaths, marriages), National Health Service Central Register (NHSCR) data (migration in or out of Scotland), NHS data (cancer registrations and hospital discharges), and education data (Hattersley & Boyle 2007; Boyle et al. 2009).
Geographical data are an integrated part of the SLS. All SLS data are coded onto a number of different geographies, including census, health and other administrative areas, at a range of spatial scales. These geographies relate to the geography existing at the time of the census, or the occurrence of a demographic event (e.g. birth). Geographies and ecological variables are essential not only for research on migration, but also for research on impacts of environmental and socio-economic contexts on individuals’ well-being. This working paper will introduce geographical data including geographical identifiers and ecological variables in the SLS.
Download output document: SLS Technical Working Paper 5 (PDF 1.5MB)
Education data available within the Scottish Longitudinal Study
Raab, G. (2013) SLS Technical Working Paper 6. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 10 September 2013. [SLS]
This document summarises the education data for school years 2007 to 2010 made available to the Scottish Longitudinal Study (SLS) from ScotXed. Such data consist of records from the School Census carried out in September of every year, attendance data obtained at the start of the following year and attainment data from the Scottish Qualifications Authority (SQA) for qualifications obtained during the year. Linkage rates from the School Census to the SLS and from the SLS to the School Census are calculated. Recommendations are made to attempt to improve the scope and quality of the data for this and subsequent years.
Download output document: SLS Technical Working Paper 6 (PDF 1MB)
The Scottish Longitudinal Study 1936 Birth Cohort
Huang, Z., Feng, Z., Dibben, C., Brett, C. & Deary, I. (2017) SLS Technical Working Paper 7. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 30 March 2017. [SLS]
This paper describes the creation of the Scottish Longitudinal Study (SLS) birth Cohort of 1936 (SLSBC1936). It is structured around the existing SLS. We took the SLS birth date sample from the Scottish Mental Survey of 1947 (SMS1947, a cognitive ability test that included almost all Scottish children born in 1936) and linked it to the 1939 National Register, the National Health Service Central Register (NHSCR) and the SLS. The outcome of the project is a powerful life course dataset containing information from childhood to old age.
Download output document: SLS Technical Working Paper 7 (PDF 5MB)
RESEARCH WORKING PAPERS
Movement from ill health related economic inactivity into employment and its impact on health: evidence from the Scottish Longitudinal Study
Popham, F. & Bambra, C. (2008) SLS Research Working Paper 1. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 28 April 2008. [SLS]
Using data from Scotland, a country with high rates of ill health and economic inactivity, this study investigated the likelihood of movement from health related economic inactivity into employment, and whether this was associated with demographic characteristics or socio-economic status. It also examined whether those who had moved into employment were more or less likely to report a longstanding limiting illness. ...
Download output document: SLS Research Working Paper 1 (PDF 88KB)
Unemployment, mortality and the problem of health-related selection: Evidence from the Scottish Longitudinal Study
Clemens, T., Boyle, P. & Popham, F. (2009) SLS Research Working Paper 2. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 17 February 2009. [SLS]
Associations between unemployment and mortality are often complicated by processes of health-related selection. Testing whether unemployment causes health deterioration is complicated because failing health may increase the probability of unemployment. In some previous studies of unemployment and mortality a ‘wear-off’ period, after employment status is observed, is used which ignores the first few years of mortality events. It is assumed that selection effects will wear-off during this period. In this study we aim to test the effectiveness of using wear-off periods.
The effect of neighbourhood housing tenure mix on labour market outcomes: A longitudinal perspective
Van Ham, M. & Manley, D. (2009) SLS Research Working Paper 3. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 14 April 2009. [SLS]
This paper investigates the effect of different levels of neighbourhood housing tenure mix on transitions from unemployment to employment and the probability of staying in employment for those with a job. We used individual level data from the Scottish Longitudinal Study (SLS), a 5.3% sample of the Scottish population, covering a 10 year period. We found a strong negative correlation between living in deprived neighbourhoods and labour market outcomes (getting or keeping a job). We found a small, but significant, positive correlation between living in mixed tenure (40-80% social housing) streets and transitions from unemployment to employment. In the conclusion we discuss the extent to which we think these results can be interpreted as ‘neighbourhood effects’ or selection effects.
Does widowhood increase mortality risk? Comparing different causes of spousal death to test for selection effects
Boyle, P., Feng, Z. & Raab, G. (2009) SLS Research Working Paper 4. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 27 November 2009. [SLS]
We consider the effect of spousal bereavement on mortality by different types of spousal death. We expect some causes of death to be correlated with socioeconomic characteristics and others not to be. Equality in the ‘bereavement effect’ across different causes of death would suggest a causal effect of widowhood, while no bereavement effect for uncorrelated causes of death would suggest that selection effects have a role.
Selective internal migration. Does it explain Glasgow’s worsening mortality record?
Popham, F., Boyle, P., O'Reilly, D. & Leyland, A. (2009) SLS Research Working Paper 5. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 16 December 2009. [SLS]
Glasgow is one the least healthy cities in Europe. The mortality difference between Glasgow and the rest of Scotland has been increasing and mortality rates are higher than Glasgow’s excess deprivation would suggest (the ‘Glasgow Effect’). One plausible explanation for this excess is selective migration. ...
Post-partum psychosis in Scotland 1991 to 2006
Raab, G. & Barbour, R. (2010) SLS Research Working Paper 6. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 14 January 2010. [SLS]
This study has investigated post-partum psychosis by linking hospital admissions with a psychotic diagnosis in the postnatal period to women members of the Scottish Longitudinal Study who gave birth between the 1991 census and the end of follow-up (2007). Admission rates were found to be towards the lower end of what has been found in comparable studies elsewhere, mainly in Scandinavia. The overall admission rate in the 3 months after a birth was 0.47 per 1000 births (95% confidence interval 0.29 to 0.74). Rates were much higher in those with a previous psychotic hospital admission at 1 in 10 births in the 3 months post- partum and 1 in 5 in the nine months post-partum. Those with a previous psychotic hospitalisation accounted for over 70% of all cases identified. There was no evidence that rates varied by parity or by the age of the mother, but increased rates were found for births that were registered by the mother alone rather than jointly by both parents, for mothers from the most deprived areas and those from households where the head could not have a social class assigned, most often because they had never worked.
Inter-sectarian couples in the 2001 census
Holligan, C. & Raab, G. (2010) SLS Research Working Paper 7. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 15 January 2010. [SLS]
We have used data on over 111 thousand couples where both partners were born in Scotland, aged 16 to 74 and with one member an SLS member, who reported in the 2001 census that they were raised in a Christian denomination, or that they were raised with no religion. This sample represents approximately 11% of all such couples in the census. ...
The proportion of inter-sectarian (RC and protestant) couples has increased steeply for the youngest age groups and in the West of Scotland they make up around 25% of all couples. ... Those in a religiously mixed partnership are more likely to have no current religious practice, but RCs in mixed partnerships are the most likely to maintain their religious practice of upbringing.
Taken together these findings suggest a breakdown of sectarianism in Scotland between RC and others. This is accompanied with an increase in secularism and with some evidence of the separation of those with no religious upbringing, who are generally a disadvantaged group, from others.
Migration and Religion in Scotland: A study on the influence of religion on migration behaviour
Van Dijck, J., Feijten, P. & Boyle, P. (2010) SLS Research Working Paper 8. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 1 November 2010. [SLS]
This paper analyzes the influence of individual religion on internal migration in Scotland. Two aspects of religion are studied: denomination and secularization. Drawing on data from the Scottish Longitudinal Study, a 5.3% population sample linking the 1991 and 2001 censuses, this paper uses the location- specific capital theory to argue that religious individuals are less likely to migrate than non-religious individuals and that Catholics are less likely to migrate than Protestants. It also uses the modernization theory to argue that individuals who moved from a rural to an urban area are more likely to have become secular. The findings corroborate the hypotheses and thus confirm that individual religion is still an important factor in explaining contemporary internal migration.
Download output document: SLS Research Working Paper 8 (PDF 131KB)
Teenage mothers and fathers in Scotland 1991 to 2001
Raab, G. & Henderson, M. (2010) SLS Research Working Paper 9. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 1 November 2010. [SLS]
This report investigates the factors that predict young parenthood in young men and women and examines the consequences of being a young parent in the years following the birth. The Scottish Longitudinal Study (SLS) holds 1991 Census data for a 5.3% sample of the Scottish population. Young people who became mothers and fathers between 1991 and 2001 were identified from linked birth records. The 2001 Census data was used to identify how young parents differed from other young people in terms of their socio- economic status, education and health. The sample studied included 2,282 teenage mothers and 811 teenage fathers. The reason for the smaller number of fathers was, to a large extent, due to the fact that the fathers of young mothers’ babies tend to be older and the same age range was used to identify each sex. The use of census data which is completed by a very high proportion of the population means that our study will be much less affected by participation bias which can affect other surveys of young parents. ...
Estimating an occupational based wage in the census: a mixed model approach to generate empirical bayes estimates
Dibben, C. & Clemens, T. (2012) SLS Research Working Paper 10. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 9 October 2012. [SLS]
Commonly, a lack of income information in the census is mitigated analytically through the use of proxy measures. Area based deprivation scores are the most common of these in which various domains of deprivation, obtained from the census, are aggregated to different spatial scales to give an indication of the characteristics of an area. Furthermore, other measures of socio-economic position such as occupational social class and education are often used to proxy some of the effect of material and financial circumstances. Though both measure independent components of socio-economic confounding (SEP) they may not capture entirely the health effects of income (Galobardes, Shaw et al. 2006). An alternative approach is to produce an estimated synthetic measure using regression modelling. However, whilst such a technique has been utilised for the production of aggregated area estimates of income (Williamson and Voas 2000), the potential for estimates generated at the individual scale, within the UK census, have yet to be investigated. The aim of this working paper is to investigate the usefulness of multilevel models based on Standard Occupation Classification (SOC) groups to estimate an occupational based wage measure in the census.
Download output document: SLS Research Working Paper 10 (PDF 392KB)
Flexible ageing: new ways to measure the diverse experience of population aging in Scotland, using the Scottish Longitudinal Study
Spijker, J. & MacInnes, J. (2015) SLS Research Working Paper 11. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 5 November 2015. [SLS]
...This project takes a sociodemographic approach to assess the challenges and opportunities of ageing populations in different contexts. Single-year of age and sex population estimate data and life expectancy data from the National Records of Scotland (NRS) (formerly General Register Office for Scotland (GROS)) ―for the period 1981-2011― and the Scottish Longitudinal Study (SLS) ―related to the 1991 and 2001 censuses ― are used to construct new measures of population ageing based on years of remaining life expectancy rather than years since birth. In other words, we treat age in terms of years left until death rather than calendar age. While the latter is routinely used in the social sciences as well as in public policy (e.g. for the calculation of dependency ratios) because of the straightforward availability of data and their relevance to eligibility criteria set by public policy for e.g. pension entitlement and other social benefits, these measures do not consider the impact of one of the two main drivers of population ageing: improvements in survival, which has been responsible for large gains in life expectancy among older ages over the last half a century in the case of men and an additional several decades for women. While the NRS data allows for the analysis of a longer time period, the SLS data permits a more detailed analysis of the level of ageing, for instance by making estimating for specific sub-populations including different categories of (former) occupation, marital status, subjective rated health and level of area deprivation. The SLS data also allow us to estimate the size and composition of the elderly according to these census variables and produce dependency ratios of the older population to the employed rather than the usual working-age population in Scotland. The goal then is to see if these tell a different story from the conventional old-age dependency ratios.
Download output document: SLS Research Working Paper 11 (PDF 10MB)
The relationship between educational outcomes and family socioeconomic position in Scotland: the role of low birthweight and child development
Playford, C.J., Dibben, C. & Williamson, L. (2017) SLS Research Working Paper 12. Longitudinal Studies Centre Scotland: Edinburgh/St Andrews, 23 May 2017. [SLS]
There is a long-standing association between family socioeconomic position and educational outcomes. This study seeks to explore whether variation in low birthweight and child development by family background explain these differences. The relationship between socioeconomic position, birth outcomes and pre-school child development is complex and requires an interdisciplinary assessment. Using linked administrative data from Scotland, our findings suggest that birthweight and child development partially explain this variation but cannot entirely account for socioeconomic differences in educational outcomes.
Between the NEET and the tidy – Exploring ‘middle’ outcomes in Scottish school qualifications
Gayle, V., Playford, C., Connelly, R. & Murray, S. (2016) CPC Working Papers, 76, 1 March 2016. ISSN: 2042-4116 [SLS]
Despite changes in the education system the qualifications that are gained at school remain important for young people’s pathways and trajectories. This paper is an element of a wider on-going programme of theoretically informed empirical analyses, which examine young people’s educational outcomes. The empirical work is situated within an overarching theoretical sociological framework which focuses on the outcomes of ‘ordinary’ young people who are neither educationally unengaged, nor part of an educational elite. In this phase of the work we focus on outcomes in Scottish school-level qualifications.
This research is original in that it uses administrative data from the Scottish Qualifications Authority that is held as part of the Scottish Longitudinal Study. We begin by reconsidering challenging issues associated with measuring school -level qualifications. To address these challenges we undertake analyses of pupil’s subject-area outcomes using a latent variable modelling approach. A novel aspect of the work is that we undertake a sensitivity analysis to compare a standard technique for latent group assignment (modal assignment) with a recently proposed alternative (proportional assignment).
The overall message is dispiriting because after half a century of comprehensive education in Scotland, school outcomes remain stratified both by gender and by a pupil’s social background. The analyses uncovered four main latent educational groups. One group had very positive outcomes and pupils in this group were generally more socially advantaged. By contrast another group had very poor outcomes and pupils in this group were generally more socially disadvantaged. There were two ‘middle’ groups, which both had moderate overall school Standard Grade outcomes. These two ‘middle’ groups were similar in their overall outcomes, but at the subject area-level their outcomes were notably different. One group were more likely to gain a Credit pass in English, but were relatively less likely to gain Credit passes in Mathematics and Sciences. The other group were unlikely to gain Credit passes in English and Mathematics, but were more likely to gain Credit passes in Sciences. These pupils with ‘middle’ or ‘moderate’ outcomes in school Standard Grades are a sociologically important group that should not be overlooked.
The latent variable approach offers an informative set of typologies that are likely to be impactful because they can be used to better understand patterns of educational outcomes. These typologies are important because they can directly inform current debates on raising standards in Scottish schools.