In Global Context: Mental Health Study
Indiana University . Bloomington . Indiana
Spain Sample Design
National = Mainland, and the
Sampling spots: 133 Municipalities.
Number of routes: 155.
Type of interview: At respondent’s home.
Sampling bias: ± 2.88 at most.
Sample stratification: The selection criteria was that of proportional distribution of
interviews according to population in each of the 17 Autonomous Regions
and according to community size.
Those municipalities with more than 500,000 inhabitants
are of compulsory selection; the rest come out from a random draw.
The same number of interviews per Autonomous Region is
kept in each study. The only Autonomous Regions that always have
interviews at all levels of community size are: Andalucía, Cataluña and
País Valenciano. The strata of the other regions are drawn each month.
of interviews that proportionally corresponds to each province and type of
community is also taken into consideration within each Autonomous Region.
Moreover, we keep track of the population settlements
that are selected so that they are not repeated, and in large
cities we distribute the interviews among the districts.
Household selection: once the number of interviews to be done has been established (by size
of community and Autonomous Region), we use a computerised system to
randomly extract municipalities and electoral sections within them. The
number of electoral sections randomly selected is related to the total
number of interviews to carry out in the municipality.
A random route system is applied for household
selection; in all the buildings that the interviewer passes by (on either
side of the street, depending on the side where the starting point is),
one every three housings is selected; in case of refusal or non-contact,
the interviewer goes to the next household.
Respondents’ selection: age and sex quotas are used for the respondent’s selection. These
quotas are established in each sampling point according to cross
tabulation between size of community and age and sex at national level and
also at regional level in the three regions above mentioned (Cataluña,
Andalucía, País Valenciano). In the other regions the distribution is
· Achieved sample: the distribution of the 1,217 interviews of this study (1,202 = achieved sample) is shown on the next table. The figure on the left in each cell corresponds to the desired sample and the figure on the right corresponds to the achieved sample.
Training for interviewers: Personal training for the interviewers was conducted
100% of the questionnaires were cleansed.
Telephone supervision of 280 interviews (23%), ASEP.
15 questionnaires were cancelled as results of the cleansing/ supervision
1022 E. Third Street, Bloomington, IN 47405 (812) 855-3841