Growing Up in Scotland (GUS) data workshops
Growing up in Scotland is a longitudinal research study tracking the lives of thousands of children and their families from the early years, through childhood and beyond. The main aim of the study is to provide new information to support policy-making in Scotland but it is also intended to provide a resource for practitioners, academics, the voluntary sector and parents. It studies two birth cohorts and their parents, with 8 sweeps of the first birth cohort already completed or in the field. This makes good skills in data management and manipulation a prerequisite for using the data effectively, as well as knowledge of how to deal with panel or repeated measures designs in analysis of the data.
Dr Valeria Skafida, Centre for Population Health Sciences, The University of Edinburgh,
Dr Morag Treanor, Centre for Population Health Sciences, The University of Edinburgh,
This event, organised by AQMeN in collaboration with the Centre for Research on Families and Relationships (CRFR) and ScotCen was run in three parts:
17th March 09:30 - 12:30
Part 1: Introduction to Growing Up in Scotland
This seminar provided an introduction to the Growing Up in Scotland study and covered topics such as: survey design and methodology including sampling; weighting and response; and content - including topics, themes and standard measures. The seminar also covered how to access GUS data and documentation.
17th March 13:30 - 17:00
Part 2: Workshop 1
The first workshop described the survey and its organisation and reviewed the data management skills necessary to organise the data from the various GUS datasets into formats suitable for analysis. The workshop used STATA. Previous experience of using this esoftware package was of great advantage, but was not a pre-requisite of the workshop. Whilst it included some basic STATA training for those more familiar with SPSS, those without previous experience of STATA would already have a good understanding of data management (merging and restructuring files) and preferably some experience of dealing with longitudinal data in SPSS.
24th March 13:30 - 17:00
Part 3: Workshop 2
The second workshop looked at the special challenges which arise when dealing with panel, cohort or repeated measures data. Such data has the great advantage that causal processes can be easier to identify because variables on the same case are measured at different time points. However some standard statistical procedures are inappropriate since measures taken on the same subject at different time points will not be independent of each other. There are other challenges that must also be dealt with: sample attrition and weighting are more complex issues than in repeated cross sectional studies and care must be taken to correctly identify the target population to which any analysis refers. Participants were required to have a good knowledge of at least one statistical processing package: up to the ability to run and interpret multiple regression analyses. Knowledge of STATA or familiarity with GUS was an advantage.