Centre for Longitudinal Studies



* In the case of NCDS pre-March 91 birth events, though, we want 'pregnum' to be zero,
* as we're deleting all reference to that event.

If (pre91flg=1 and key(#i)=40)xpregnu(#i)=0.

end if.

end loop.

* Copy the 'pregnancy' variables associated with that baby to the correct pregnancy slot (1-8).

* Note that within the 3-loop structure (#i,#j,#k) the copying is independent of the value of #k.

* i.e. anything up to five babies wrongly bunched in one pregnancy will have identical

* pregnancy data copied to different new pregnancy slots.

* Note also that, in the case of NCDS pre-March 91 birth events, the data for the 'eighth'
pregnancy

* (i.e. that corresponding to the 40th birth event) will be copied, and in all such NCDS cases
we know this is blank.

loop #i=1 to 8.

loop #j=1 to 8.

loop #k=1 to 5.

Do if (key(#i)=(5*#j+#k-5)).

compute xpregj(#i)=vpregj(#j).

compute xcg1prg(#i)=vcg1prg(#j).

compute xcg2prg(#i)=vcg2prg(#j).

compute xcg3prg(#i)=vcg3prg(#j).

compute xcpregb(#i)=vcpregb(#j).

compute xpregk(#i)=vpregk(#j).

compute xpregkw(#i)=vpregkw(#j).

compute xpregl(#i)=vpregl(#j).

end if.

end loop.

end loop.

end loop.

end if.

* Now address every single case that needed a manual alteration.

Do if (nserial='083034N').

Compute changed=2.

Compute key40(1)=21.

Compute key40(6)=11.

Compute key40(7)=16.

Compute key40(11)=6.

Compute key40(16)=1.

Compute keypreg(1)=5.

Compute keypreg(2)=3.

Compute keypreg(3)=2.

Compute keypreg(4)=1.

End if.

* The following cases all had exclusively pre-March 1991 birth event data, and were

* interviewed at NCDS5, so delete all birthslot information (i.e. make changed=2, but no key
values).

Do if (nserial='094029H' or nserial='330027Q' or nserial='986039J').



More intriguing information

1. The name is absent
2. The name is absent
3. Valuing Farm Financial Information
4. Are combination forecasts of S&P 500 volatility statistically superior?
5. Inhimillinen pääoma ja palkat Suomessa: Paluu perusmalliin
6. The name is absent
7. The name is absent
8. A Classical Probabilistic Computer Model of Consciousness
9. On s-additive robust representation of convex risk measures for unbounded financial positions in the presence of uncertainty about the market model
10. The name is absent