Tracking Family Structure in ALSPAC

1
Tracking Family Structure in ALSPAC
David W Lawson Ruth MaceHuman Evolutionary
Ecology GroupUniversity College London
2
Tracking Family Structure in ALSPAC
1. Family Structure Data in ALSPAC
- Fathers - Siblings
2. An Example of Using the Data
Lawson, D.W. Mace, R. (submitted). Trade-offs
in modern parenting a longitudinal study of
sibling competition for parental care. Evolution
and Human Behavior
3
Distribution of Family Data in ALSPAC
400 variables reference family structure. Most
data unevenly distributed across 6/7 key
points (mother questionnaires)
(Child questionnaires) With other relevant data
spread in between.
4
Tracking Fathers A. Parental Relationship at
Birth
i. Mothers Relationship to Childs Father in
Pregnancy
(Mothers Questionnaire A a520, a521)
75 married and cohabiting 16 unmarried
cohabiting 5 unmarried not cohabiting
5
Tracking Fathers A. Parental Relationship at
Birth
ii. Prior Length of Relationship to Childs
Father at Pregnancy (Mothers Questionnaire A
a523yy)
Mode 3 years
6
Tracking Fathers B. Father Presence Over Time
G
H
J
M
Q
A
i. Option 1 Age at which biological father
leaves
20 of absent fathers, absent from birth.
(Key Point Data g521, h401, j375, Mc4b1yr,
Mc4b1mth, QC4bi)
Advantages Gives us exact age dad
left. Disadvantages Doesnt tell us about
replacement father figures.
7
Tracking Fathers B. Father Presence Over Time
G
H
J
M
Q
A
ii. Option 2 Live-in Father-Figure Variable x
Cohab data
(Key Point Data h400 ,h376 , j374, j365b, Mc4a,
Mc1b2, g520, g506, g521, h401, j375, Mc4b1yr,
Mc4b1mth, QC4b1 QC4a, QC1b2) Objectifies
father-figure as live-in partner of mother
Advantages Tracks non biological father
figures. Can impute some missing data. Necessary
for coding sibling relatedness Disadvantages Res
tricted to unevenly distributed key
points Note Able to cross reference each option
and impute some missing.
8
Tracking Fathers C. Investment from Absent
Fathers
G
H
J
M
Q
A
i. Financial Investment How frequently does the
bio father provide financial support?
(Key Point Data g523, h403, j377, Mc4b3, QC4b3)
ii. Time Investment How often does the bio
father see the child?
(Key Point Data g522, h402, j376, Mc4b2, QC4b2)
9
Tracking Fathers C. Investment from Absent
Fathers
Absent fathers more likely to invest later in
study period
10
Tracking Fathers D. Relationship Status Over Time
i. Option 1 Matching Repeated Questions on
Cohabitation and Marital Status
Many repeated but differently distributed
questions, could be matched up Is mum
cohabiting with partner?, current marital
status? Disadvantages Difficult tell if partner
is changing even if mothers status stays the
same. Does not tell us about fathers
relationship to the study child.
ii. Option 2 One-off Relationship Survey at
Questionnaire M (7y1m)
M
Advantages Great Survey No. of Live in
Relationships of Mother, Dates for start and end
for last 4 relationships. Disadvantages Only
available for the 8,000 that filled in
questionnaire M. No comparable surveys in later
waves, stops at M.
11
Tracking Siblings A. Number of Siblings
i. Total Number of Resident Siblings
(Key Point Data g530, g531, g532, h411, h410,
h412, j382, j383, j384, Mc5a, Mc5b, Mc5c, QC5a,
QC5b, QC5c)
ii. Total Number of Non-Resident Siblings
(Key Point Data g536, g538, g540, h415, h417,
h419, j386b, j387b, 388b, Mc6anum,Mc6bnum,
Mc6cnum, QC6ano, QC6bno, QC6cno)
Limitations Residence is not clearly objectified
e.g. does it include kids at boarding
school/university? Or kids living with different
carers??
At 10 years, 10 of families have at least one
non resident sibling
12
Tracking Siblings A. Number of Siblings
iii. Sum Total Number of Siblings
Around 50 of kids are only children at birth,
but only 10 are only children by Age 10.
13
Tracking Siblings B. Relatedness
i. Number of Siblings Related to Live-in Father
Figure
(Key Point Data g530, g531, g532, h411, h410,
h412, j382, j383, j384, Mc5a, Mc5b, Mc5c, QC5a,
QC5b, QC5c) Data on who the current father
figure is, so we can subtract study child from
total number and also know if a man is actually
in the house!
Limitations All of this data contains no
information on age or sex of children
14
Tracking Siblings C. Sex Relative Age

• Centring Relative Age and Sex Data to Key Points

(kb590, kb591, kb592, kd451, kd452, kd453, kd432,
kd453, kd454, j420,j422, j4228a/b j420a/b,
j422a/b, kq637, kq636a/b kq630, kq631, kq633,
kq634, kq632, kq635, kud63b1bro, kud6b1sis,
kud6b2bro, kud6b2sis) Lots of imputation...
15
Tracking Siblings C. Sex Relative Age
Limitations All of this data specifically makes
reference only to currently coresident
children. Imputation leads to some
inconsistencies in data Still dont how actual
age of siblings. YSibs just arrival
windows OSibs only know DOB of next oldest from
interpreg interval.
16
Tracking Siblings D. D.O.B.
i. Data from past relationship survey at
Questionnaire M (7y 1m)
M
Advantages Good Survey DOB of each child
associated with last 4 live in relationships
Disadvantages Only available for the 8,000
that filled in questionnaire M. Dont have
comparable surveys in later waves, stops at
M. While tells us relatedness and DOB, doesnt
tell us sex or residence of each child.
17
Acknowledgements
Jon Heron, Jean Golding ALSPAC (Data)
Jill Greenberg(Images)
ESRC Parkes Foundation(Funding)
Human Evolutionary Ecology Group
Lawson, D.W. Mace, R. (submitted). Trade-offs
in modern parenting a longitudinal study of
sibling competition for parental care. Evolution
and Human BehaviorLawson, D.W. Mace, R. (in
press). Sibling configuration and childhood
growth in contemporary British families.
International Journal of Epidemiology