Professional Anglo-Indians and those in politics were all
included in the 4-6 vertical segmentation category. (37)
An East Indian school friend in Bandra, Bombay who had
attended an Anglo-Indian school with the researcher
(1944-1956) volunteered her services as the coordinator in
India. All correspondence from respondents written after
April 1990 was directed to her. She possessed excellent
language skills in English, Hindi and Marathi and was a
well known social worker who was deeply interested in the
research. Bombay was selected as the first city, because
the researcher was familiar with the educational system in
Bombay and had been invited by the coordinator to use her
home as a working base.
The next section describes the cluster sampling of the
target population.
6. CLUSTER SJkMPLING CONSTRAINT The target population was
widely dispersed. It would have been impractical to select
the respondents at random because an inordinate amount of
time would have to be spent travelling about to meet them.
A second consideration was the airline ticket for travel in
India. It had to be used within four weeks in a clockwise
or anti-clockwise direction, that is, one could not return
to the same city, unless it was the starting point of the
journey. Hence, there was no alternative to using a
cluster sample. There were two stages in the process of
multi-stage cluster sampling which were used. (38)
Stage One: Selection of a sample of Anglo-Indian leaders of
the community, which included Members of Parliament,
ex-Members of Parliament, Members of the State Legislative
Assemblies, ex-Members of the Legislative Assemblies.
Stage Two: Selection of a sample of schools using the
182