Send the link below via email or IMCopy
Present to your audienceStart remote presentation
- Invited audience members will follow you as you navigate and present
- People invited to a presentation do not need a Prezi account
- This link expires 10 minutes after you close the presentation
- A maximum of 30 users can follow your presentation
- Learn more about this feature in our knowledge base article
Do you really want to delete this prezi?
Neither you, nor the coeditors you shared it with will be able to recover it again.
Make your likes visible on Facebook?
Connect your Facebook account to Prezi and let your likes appear on your timeline.
You can change this under Settings & Account at any time.
CALLS Hub Prezi
Transcript of CALLS Hub Prezi
Census data forms the core of each of the 3 LSs. In the UK, it is compulsory for all households to complete a census form every 10 years. Because of this legal requirement, the LS datasets are far less prone to loss of members than most other data sources and panel surveys. This allows us to follow up individuals across time much more effectively.
Each census provides us with a wealth of information on such things as:
The ONS LS has captured census data from 1971 onwards, whilst the SLS and NILS hold data from 1991 onwards. All three studies have been expanded this year to include data from Census 2011.
In order to create each LS, random birthdates are used to select individuals from the first census in the study, and these become the sample members. Data on them and the members of their household are collected from the census. Sample members are then picked up at the next census 10 years later, and a new wave of data is collected. Data for household members is not linked through from census to census, but instead held separately at each point.
There are several ways in which LS members may be lost from the samples - eg, if they migrate to another part of the UK or emigrate, if they die, or sometimes simply the data needed to make a linkage is not available. In a similar fashion, at each census new LS members with one of the selected birthdates are added to the datasets. These new members may have been born since the last census, or may have immigrated into the country.
But census data is just the beginning of the story...
Other data sources
As well as census data, the LSs contain information from a variety of other administrative data sources. This data is then linked in to the core census data for the LS sample.
For example, information from registrations of births and deaths are contained in all three LSs, and Scotland also has marriage registration events.
The LSs each have unique linkages to other datasets, such as education (SLS), health data (SLS, NILS), and property datasets (NILS). Whilst some data is held within the LS - eg, cancer registrations in the ONS LS - other data is linked in on a project-by-project basis - eg, health data in the SLS.
Welcome to our brief introduction
to CALLS Hub and the LSs
CALLS is the Census & Administrative data LongitudinaL Studies Hub. We are an ESRC-funded project and were set up with the main objectives of supporting, promoting and harmonising the work of the three UK Longitudinal Study (LS) Research Support Units (RSUs).
This Prezi will tell you a little more about it...
As we have seen, there are various types of data linked into the LSs.
Perhaps the largest linkage is the SLS linkage to NHS Scotland's comprehensive ISD datasets - http://www.isdscotland.org. This contains information on maternity, mental health, cancer, deaths and other health data, providing a rich research resource.
Other linkage examples include the NILS linkage to NI Land and Property Services data, which gives detailed information on house prices and property characteristics - http://www.dfpni.gov.uk/lps/
If you have another external data source that you would like to be linked in to the LS data sample for your project you can discuss this with your Research Support Officer when developing your application.
ONS Longitudinal Study of England & Wales
The ONS LS is the oldest of the three LSs, containing data from the 1971 census onwards. With the addition of the new Census 2011 data this now allows 40 years of follow-up, presenting a real opportunity for research across the life-span. The data include linkage to cancer registrations data, births and deaths.
The ONS LS is also the largest of the LSs, with approx 950,000 members. This represents a 1% sample of the population of England & Wales.
The dataset is supported by CeLSIUS at UCL, and by ONS at Titchfield. For further information see:
Northern Ireland Longitudinal Study
The NILS boasts the largest proportional sample of the three LSs. Its sample of 500,000 members represents approximately 28% of the population of Northern Ireland giving it considerable statistical power. The NILS has recently been expanded backwards to 1991, providing 20 years followup, and in Spring 2015 will be extended right back to 1981.
There is a principal focus on health-related research when using the NILS, which might include research on migration, fertility, mortality and inequalities. The dataset contains linkages to useful data such as GP prescribing, as well as a linkage to detailed property information.
The dataset is supported at NILS-RSU, Belfast. To find out more: http://calls.ac.uk/ls-units/northern-ireland-longitudinal-study-northern-ireland/
Scottish Longitudinal Study
The Scottish Longitudinal Study is the broadest LS in terms of the breadth of data linkages it has secured. Most notable is a linkage to NHS Scotland health data, which are amongst the best health data in the world. Other linkages include weather, pollution and extensive education data, with more dataset linkages planned.
The SLS sample represents approximately 5% of the Scottish population, and has approx 274,000 members. It is created and maintained by the SLS-DSU at the Longitudinal Studies Centre Scotland, with a safe-setting in Edinburgh.
More information can be found at: http://calls.ac.uk/ls-units/scottish-longitudinal-study-scotland/
Until recently, it was not possible to perform joint analysis using data from more than one of the LSs, with researchers instead having to combine results from separate analyses post-hoc.
A new methodology has been refined by staff at the SLS-DSU, called eDatashield. This technique means that researchers can now perform GLM analyses on more than one LS as though they were part of the same dataset. This opens up the possibility of comparative research, e.g., the pilot study designed to test the eDatashield method: 'A comparative study of the relationship between deprivation and health status in Northern Ireland and Scotland' - see http://sls.lscs.ac.uk/projects/view/2011_005/
If you are interested in using more than one LS for your analysis, please contact the CALLS Hub helpdesk on email@example.com
What types of research have the LSs been used for?
LS-based research has informed academic and policy fields on a wide range of topics including health psychology, neighbourhood effects, migration and health inequalities. Some project examples include:
"Area influences on health: does the extent of community or religious segregation matter?" [NILS project 032]
"Residential location, migration and occupational achievement in Scotland 1991-2001" [SLS project 2007_005]
"Adult cancers near overhead power lines" [ONS LS project 30104]
"The growth of mixed-ethnic unions and their changing geographical distribution between 1991 and 2001" [SLS project 2007_016]
"The causal effect of schooling on social mobility: findings from a natural experiment" [ONS LS project 30139]
"Social Harm and the Elderly in Northern Ireland" [NILS project 027]
You can explore outputs from LS-based research at:
The SYLLS project - Synthetic Data Estimation for UK Longitudinal Studies
The LSs are rich and valuable data sources, but due to their sensitive nature, access is restricted to approved researchers in secure safe-settings. Being unable to download data to test and develop analyses has historically been a barrier to use of the LSs.
The SYLLS project was set up to overcome this problem by developing two complementary methods for generating synthetic data products:
statistical modelling with conditional specification is used to generate bespoke synthetic datasets for individual research projects; and
microsimulation is used to generate synthetic longitudinal data 'spines' for most frequently used variables in each of the national longitudinal studies.
This synthetic data will allow potential users to explore likely patterns in the real data to develop their hypotheses and syntax prior to booking a visit to a safe-setting. The data may also be used in academic teaching settings as well as LS training workshops.
The technique developed for generating synthetic versions of real datasets is now available as the R package 'synthpop' which can be found at http://cran.r-project.org/web/packages/synthpop/index.html
You can find out more about the work of SYLLS at:
CALLS Hub - a portal to the LSs
One of the main functions of CALLS Hub is to act as a portal to the three UK LSs. This is partly achieved by our website, which provides:
information about the LSs and RSUs
the most comprehensive collection of relevant census forms available anywhere online
a searchable database of all outputs from the three LSs
a data dictionary which allows searches for variables across all three studies - particularly useful if you are considering using more than one LS for your work
a central point to find out the latest news from the LSs
Getting in touch...
You can find out more about CALLS Hub and the LSs on our website or by following us on Twitter.
We also provide a helpdesk for all LS-related queries.
tel: 01334 464021