The Longitudinal Study of Young People in England (LSYPE) Clare Baker David Simpson Helen Wood
Mar 28, 2015
The Longitudinal Study of Young People in England (LSYPE) Clare BakerDavid SimpsonHelen Wood
Agenda for the morning
An overview of LSYPE (45 minutes to 1 hour) Comfort Break (10 minutes) An introduction to iLSYPE (45 minutes to 1 hour) Comfort Break (10 minutes) A chance to use iLSYPE (30 minutes) End (1pm at the latest)
Presentation OverviewWhat is LSYPE? Who takes part Survey content Data collection Linked data
Benefits
Uses of LSYPE
Data Enhancement and Access to LSYPE data
Challenges and the future of LSYPE
What is LSYPE?
A unique and invaluable dataset for Government analysis
and policy making
Initially funded by HM Treasury and intended as a cross-government project – currently managed by DCSF
Research aims –
factors affecting educational progress
transitions following compulsory education,
effects of background and behaviours,
attitudes and experiences
Not just restricted to education
Our unique selling point!
What is LSYPE? – Why Longitudinal? A longitudinal design provides more insight into causal
processes than cross-sectional studies. LSYPE can investigate:
– whether aspirations had been lower from a young age for certain groups and track these over time
– factors associated with changing aspirations such as changed family circumstances, parental attitudes, exam results etc.
– Investigate the degree to which intentions are converted into applications and acceptances for different groups.
What is LSYPE?
Longitudinal in design - the same young people interviewed every year from age 13/14 (2004)
Each annual LSYPE survey is referred to as a ‘Wave’. So far:
W1 Age 13/14 Year 9 2004 Data released to the public
W2 Age 14/15 Year 10 2005
W3 Age 15/16 Year 11 2006
W4 Age 16/17 Year 12 2007
W5 Age 17/18 Year 13 2008 Available on request (publication 2010)
W6 Age 18/19 Year 14 2009 Fieldwork nearly completed
Who takes part in LSYPE?Wave 1 (13/14) Young people and
both parents≈ 15,500 responding households (74%)
Wave 2 (14/15) Young people and both parents
≈ 13,500 responding households (86%)
Wave 3 (15/16) Young people and both parents
≈ 12,500 responding households (92%)
Wave 4 (16/17) Young people and one parent
≈ 11,500 responding households (92%)
Wave 5 (17/18) Young people only ≈ 10,500 responses in Wave 5 (89%)
Sampling Maintained Schools Independent Schools/PRUs
School sampling(33,000pupils)
Drawn from School Level Annual School Census (SLASC).
Oversampled deprived areas (1.5 times)Selection Probability based on Y9 pupils
Drawn from Pupil Level Annual School Census (PLASC).
Stratified by boarding status, attainment, gender.
Pupil sampling(21,000 pupils)
Drawn from PLASC. Contact details provided by co-operating schools.
Selection Probability based on Y9 pupils from Ethnic Minorities and school selection probibility
Pupils sampled from school rolls with specialist software
33/34 pupils selected at each school, or all pupils if school <34 but >5 Y9 pupils.
AchievedSample(15,500pupils)
21,000 details issued to field, of which 15,500 yielded responding households
Each year, sample issued from respondents to previous wave.
At Wave 4 an ethnic minority "Boost" sample to increase numbers.
How do we collect LSYPE data?
Waves 1 – 4: All interviews took place face-to-face.
Waves 5 and 6 – A range of modes was introduced:
on-line survey – first chance (around half) telephone survey – second chance (around one third) face to face survey – last chance (the remainder)
Participation is voluntary – respondents can drop out of the study at any time.
As a thank-you for taking part, respondents receive a High Street Voucher.
Survey Content“Main” Parent “Second”
ParentYoung Person
Employment Histories – including incomeRelationship histories (household composition)Demographics (ethnicity, gender, religion, disability etc)
Educational experiences/attainment- including bullying and truancyAspirationsOut of school activities- both positive & riskyHealth, disability & SENCaring responsibilitiesRelationship with parentsAttitudes & experiences re: local communityRelationships, sexuality etc.
Aspirations for the young personAttitudes to schooling and educationContact with servicesFamily Cohesion
Data LinkingLSYPE data can be linked to a number of other data sources to
supplement our data:
National Pupil Database (NPD) – exam results, school level information, geographic markers.
Individualised Learner Record (ILR) – Post-16 education, FE courses– Respondents were told we would link the data and asked if they
wanted to opt out. Few did.
Work and Pensions Longitudinal Study (WPLS) – DWP benefit and employment programme data as well as employment records from HMRC.
– Respondents were asked if we could link their data
Data LinkingYouth Cohort Study (YCS) – A long running cohort study from ages 16-19,
Cohort 1 started in 1985. Cohort 13 in progress
From Wave 4 onwards (2007) LSYPE sample members were the same age as YCS Cohort 13. LSYPE and YCS were harmonised to create an extra large sample.
0
5,000
10,000
15,000
20,000
Wave 1 Wave 2 Wave 3 LSYPEWave 4,
YCSSweep
1
LSYPEWave 5,
YCSSweep
2
Sa
mp
le S
ize
LSYPE YCS
Benefits Unique - the only study to have followed this cohort of young people
in this much detail.
Outcomes- cross sectional surveys show what is happening. Longitudinal surveys can show the consequences- useful for policy development.
Large sample size and low attrition - reduced potential for bias.
Highly engaged respondents- excellent rapport with researchers- allows us to ask more challenging questions.
Linked administrative data - of use to many government departments and policies.
Long shelf life – A strategic, multi-purpose survey. Doesn’t focus purely on specific policies.
How have LSYPE data been used?
By the Department:
Annual National Statistics publications. See: http://www.dcsf.gov.uk/rsgateway/DB/SBU/b000795/index.shtml http://www.dcsf.gov.uk/rsgateway/DB/SBU/b000850/index.shtml
Internal unpublished analyses:
RPA/NEETs:
Well qualified NEETs
Barriers to re-engagement
EMA eligibility/take-up
Jobs without training
Behavioural problems:
Contact with Educational Welfare, Social Services and the Police
Teenage Pregnancy Strategy Refreshment:
Characteristics of 17 year-old parents
REMOVE
How have LSYPE data been used?
LSYPE research framework agreement and research programme:- Fast and responsive tender process for research- Framework contractors build expertise with LSYPE
So Far:
Published
Drivers and Barriers to Educational Success
Examination Performance of Ethnic Minority pupils
In the pipeline…..
Risky Behaviours & social Activities
Disengagement from Education
Bullying
Community Cohesion
Youth Crime
Alcohol Use
IAG
REMOVE
How have LSYPE data been used?
By the Department: Annual National Statistics publications. See:
http://www.dcsf.gov.uk/rsgateway/DB/SBU/b000795/index.shtml
http://www.dcsf.gov.uk/rsgateway/DB/SBU/b000850/index.shtml Internal unpublished analyses Published analyses from a research framework
Published
Drivers and Barriers to Educational Success
Examination Performance of Ethnic Minority pupils
In the pipeline…..
Risky Behaviours & social Activities
Disengagement from Education
Bullying
Community Cohesion
Youth Crime
Alcohol Use
IAG
How have LSYPE data been used?
By other government departments: DEFRA/DCSF- rurality, attainment & post-16 participation
BIS-attitudes & intentions towards Higher Educationhttp://www.employment-studies.co.uk/pubs/report.php?id=bis_rp_003
Cabinet Office- attitudes & aspirations of those in deprived communities http://www.cabinetoffice.gov.uk/social_exclusion_task_force/short_studies/aspirations.aspx
By academic researchers: Links between income & GCSEs Smoking, alcohol & personality traits Roles of attitudes & beliefs on child outcomes
Example FindingsNumber of years students reported being bullied (from Y9 to
Y11) and their subsequent destinations post-16
70
5
78
63
8
6960
9
6762
11
64
0
10
20
30
40
50
60
70
80
90
% Full time Educationat 17
% NEET at 17 % Attained Level 2 by17
Per
cent
age
never once twice three times
Source: LSYPE Waves 1 - 5
Example FindingsPercentage of young people Not in Education Employment or
Training at 16, and at 17 by attainment in Year 11.
0
10
20
30
40
Non
e1
D-G
2 D
-G3
D-G
4 D
-G5
D-G
6 D
-G7
D-G
8 D
-G9
D-G
10 D
-G1
A*-
C2
A*-
C3
A*-
C4
A*-
C5
A*-
C6
A*-
C7
A*-
C8
A*-
C9
A*-
C10
A*-
C11
A*-
C12
A*-
C13
A*-
C14
+ A
*-C
GCSEs and equivalencies
Per
cent
age
NE
ET
NEET at 16
NEET at 17
Example FindingsAgreement with “Britain today is a place where people are
usually treated fairly no matter what background they come from”
5
5
6
15
15
14
10
13
13
42
48
53
47
49
57
61
60
62
40
38
31
32
30
25
25
21
21
12
9
9
6
6
4
4
6
4
0 20 40 60 80 100
Black Caribbean
White
Mixed
Black African
Other
Bangladeshi
Indian
Other Asian
Pakistani
Percentage
Strongly agree Agree Disagree Strongly disagree
Source: LSYPE Wave 5 and YCS 13 sweep 2
Example FindingsHow often do you speak to the following about what you
might do after Year 11?
1
1
1
2
12
13
3
4
6
15
35
41
10
12
19
34
34
33
18
22
32
32
13
9
68
61
43
17
7
4
0% 20% 40% 60% 80% 100%
Connexions
Careers Advisory Service
Teachers Outside Lessons
Teachers as part of a lesson
Friends
Family, e.g. parents and siblings
A Lot Quite A lot A Little Not Very Often Not at all
Source: Longitudinal Study of Young People in England (2005 - young people aged 14)
Example FindingsEstimated Income based EMA eligibility by activity (aged 16)
0
10
20
30
40
50
60
70
80
Ful
l-tim
eE
duca
tion
Job
with
Tra
inin
g
Job
With
out
Tra
inin
g
App
rent
ices
hip
etc. Oth
erE
d/T
rain
ing
NE
ET
All
% eligible for any EMA % eligible for £30 EMA
Example FindingsAge of Respondent’s mother by whether or not respondent has their own
child by the age of 17
27
35
22
7 1
35
26
8 48
*
25
*0
5
10
15
20
25
30
35
40
25 - 29 30 - 34 35 - 39 40 - 44 45 - 49 50 - 54 55 - 59
Age of mother when respondent was age 13/14
Per
cen
t
Do not have own children Have own children
Source LSYPE Waves 1,2 and 5Note: Weighted Sample is small for respondents w ith ow n children n=234.
Data EnhancementWaves 1-4 have been enhanced to aid users. Benefits
include:– Derived Variables– Edited, Labelled and re-ordered data– Detailed documentation– Weights to compensate for household non-response
A project currently underway to compensate for missing data:– To reduce biases due to non-response from individual
household members.– To reduce bias from missing data in Household income,
Socio-Economic Class and Parental Education.
How can I access LSYPE data?
LSYPE Datasets are available for users to download via the UK Data Archive. So far Waves 1- 4 have been deposited. See:
http://www.data-archive.ac.uk/
Wave 5 is available for users via an application to the Longitudinal Studies team and upon completion of a confidentiality agreement.
iLSYPE (more on this later)
Challenges
Making best use of longitudinal data
Resources
Lack of engagement across Government
Risks associated with changing to mixed-mode data collection
Managing expectations/demands
The future of LSYPE
Current cohort aged 18/19 and have just been interviewed for 6th time
About to start development for wave 7 interviews to take place 2010
Working closely with ESRC and OGDs for the possibility for ESRC to take over management of LSYPE after wave 7 (end 2010)
Development work for a second cohort (LSYPE2) - subject to Cross-Government funding
Any Questions?The Longitudinal Studies Team:
Clare Baker [email protected]
Helen Wood [email protected]
David Simpson [email protected]
Michael Greer [email protected]
Respondents’ website: www.nextstepsstudy.org.uk