Notes to Pages 18–35- The corresponding α- coefficients can be calculated from the β-
coefficients using the standard deviations provided in online Descrip-
tive Statistics. - In a 2- tailed test. This is of course equivalent to a 95% probabil-
ity using a 1- tailed test. Each regression uses data on all subjects who
give data on the dependent variable. Where data are missing for a right-
hand variable, we use instead the average value of all replies to that vari-
able. We also include a variable- specific dummy to reflect that the value
is missing. (Multiple Imputation methods give very similar results.) - Apart from mental health, we do not in this book examine the
effects of personality on happiness since these are largely captured in
the effects of mental health. Nor do we examine the hugely important
issue of the individual’s philosophy of life, which cannot be easily stud-
ied using large surveys, but is well covered in positive psychology and
in ancient wisdom. - We do not measure intellectual performance at 16 because the
BCS offers only a binary variable— whether the individual obtained
any O- level or equivalent grade A*– C at GCSE. - The “malaise” score.
- Emotional health remains the best predictor if we measure
child development over the whole course of childhood; see Layard,
A. E. Clark, et al. (2014). - Measured separately as the mother’s and father’s highest educa-
tional qualification at the time of the pregnancy. - Our analysis relates to any work, full- time or part- time.
- In other words we include a school dummy, and Figure 5 (b)
therefore reports the influence of the whole set of school dummies.
(This influence equals the standard deviation of the coefficients on the
school dummies.) In Chapter 14 we look separately at the effect of
primary schools on outcomes in primary schools and the effect of sec-
ondary schools, given the measured outcomes of the children when
they leave primary school.
Chapter 2. Income
- Basically a measure of real hourly wages times working hours
plus income from wealth (Becker [1964]). - See Layard (2011), chapter 9.
- http://www.cls.ioe.ac.uk/page.aspx?&sitesectionid=795&site
sectiontitle=Welcome+to+the+1970+British+Cohort+Study+(BCS70).