towards vertical mobility which otherwise would become much more
difficult?
And if so, time needed to take such degrees will be as well gender
dependent? How far will such features such as family status,
husband’s/wife’s profession and occupational status, own satisfaction with
work conditions, mobility chances, learning opportunities and alike
„.affect both decision and time to take post-graduations, for women and
for men?
The theoretical model follows the Hazard Analysis, with the usual
formalisation14:
h (t | x) = h0 (t) ex’ β ,
where x is a vector of known individual characteristics, β a vector of
unknown parameters and h0 (t) is a baseline hazard function for an
individual with x=0.
Let T represent the individual life time, with T ≥ 0 continuous, f (t)
the probability density function (pdf) and F(t) the distribution function,
such as
F(t) = Pr [T ≤ t] = ∫0tf(x) dx
Therefore, the survivor function, S(t), comes
S(t) = Pr [T ≥ t] = ∫t∞ f(x) dx
14 See, for instance, Lawless (1982) and Kachigan (1986).
16