gor
Notes from the GHS
Question
I am writing to enquire after the feasibility of using
General Household Survey (GHS) data to construct Government Office Region
(GOR) level indicators comparing the incomes of ethnic minority groups across
the English regions. Were sufficient individuals recruited from each ethnic
group in each GOR to enable meaningful and statistically significant indicators
at the GOR level to be developed? (e.g. would it be possible to compare the
usual gross weekly income of families where the household reference person
was of Indian decent living in different GORs?).
Answer
There are several complications
with this. The size of the GHS sample means that in one year the numbers in
the ethnic minority groups can be small, which is why we normally combine
3 years data when analysing ethnicity. When you are further breaking it down
by region some of the bases will be too small to be reliable. Also a new question
on ethnicity was introduced on the 2001 survey, which will only give you 2
years (2001 and 2002). And as there was no survey in 1997 and 1999, to get
3 consistent years together you would have to use 2000, 1998 and 1996.
