|
I conducted the Student Survey during fieldwork in Cape
Verde in 2000, and it provided one of the data sets for my thesis. The primary
purpose of the survey was to find out who wished to emigrate and
who wanted to stay in Cape Verde - or rather, return immediately to Cape
Verde after studying at a foregn university. I sought to explain this with
reference to various independent variables. This constituted one element in the
analysis of Cape Verdean emigration based on the aspiration/ability model which
I developed through the thesis.
The text below describes the survey itself, first the
questionnaire design and then the sampling process. Here at the Dragoeiro web
site, I have presented additional results that were not immediately relevant to the thesis,
but that are nevertheless interesting.
Questionnaire design
The principal objective of the survey was to explore the
relationship between the dependent variable, aspiration to emigrate, and a
range of independent variables that I assumed to be relevant. The first
decisive challenge was to formulate questions about aspiration to emigrate in
the best possible way. My main concern was to avoid formulations that could
have a significant affect the answer, such as 'do you wish to emigrate or not?'
With this formulation, it is easy to feel that answering no would signal
passivity or lack of initiative. What I decided, was to show interviewees a
card with different migration and non-migration options and ask which one they
preferred:
- Stay in São Vicente/Santo Antão
- Move to another island in Cape Verde
- Emigrate and stay abroad
- Emigrate and return to Cape Verde
The respondents were told that only studying abroad did not
count as 'emigration'. Those who wished to study abroad and come straight back
woud therefore be recorded as not wishing to emigrate, whereas those who wanted
to stay abroad to work for some years after university would be recorded as
wishing to emigrate. The card did not include any choice of island for those
who wished to emigrate and return, which might have been logical. However, it
was important to keep the card as simple as possible, and have the same number
of emigration and non-emigration options.
I believe that this method worked well. The most important
problems were concerned with long-term planning in itself, since many students
had vague ideas about what they wanted. Furthermore, they often had conditional
preferences of the kind 'if I get a scholarship, I wish to go to university and
come straight back, but if I don't I wish to work abroad for some years'. This
is discussed more in detail in chapter five. After chosing an option from the
card, those who wished to emigrate were asked where they wanted to go, and
which island they would prefer to live on if they were unable to emigrate.
It was important that this question came after respondents
had 'warmed up' with questions about themselves and their family, and before
any other questions about migration. Because I was anxious about influencing
the answers to this questions, I also introduced the interview saying that this
was a study about the situation of young people in Cape Verde, rather than a
study about emigration. It was true that I wanted to get a more comprehensive
view of the situation of young people, and focusing on emigration from the
outset could easily have influenced the answers. The independent variables
could be divided into five main groups:
- Demographic background variables about the respondent
and his/her parents
- Information about the household, including
socio-economic indicators
- Views and opinions about Cape Verde
- Relatives abroad, gifts and remittances
- Social integration
The intention was that some variables could be used
independently, such as gender and age, while others would be used to calculate
indices. For this purpose, it was necessary to include several questions which
shed light on the same theoretical concept from slightly different
perspectives. For instance, I wished to know if pessimism about Cape Verde's
future was positively associated with aspiration to emigrate. I then formulated
several statements and asked the respondents to say if they agreed or disagreed
with each one. In some, an affermative answer would be a sign of pessimism,
while in others, this would be the case with a negative answer. The answers
could then be analysed to see if they were consistently optimistic,
consistently pessimistic, or neither. The construction of different indices is
discussed below, in the section about analysis.
I also wished to see if those who showed a high level of
social integration were less inclined to wish to emigrate. Again, it was
necessary to include a number of questions addressing different aspects of
social integration, such as about spare time activities and the relationship
with friends.
Social status was a potentially sensitive issue, and I
decided to spread the questions related to this. Early in the questionnaire,
there were fairly uncomplicted questions about household size, the number of
rooms and utilities present. The more sensitive questions about the household's
ability to cover unforeseen expenses, and the respondents own characterization
of the household's economic situation were 'tucked' between questions about
spare time activities towards the end of the questionnaire. Such considerations
about the emotional dynamics of the question sequence are as important with a
questionnaire as with a semi-structured interview. I believe that the intervew
experience was a positive one for the vast majority of respondents. This was
important not only for the sake of the individual, but also because it affected
the attitude towards me in the student body of the schools where I conducted
interviews.
The sampling process
The survey finally included 264 interviewees sampled from a
universe of 635 students in their final (12th) year of school. The universe
included day-time students of the four state-run schools on the islands of
São Vicente and Santo Antão. There are also nocturnal classes for
older students, usually people who started working without completing secondary
education and later resumed school on an evening basis. At the time of the
fieldwork, no private schools were teaching the final year of liceu. The sample
was stratified by school and gender, so that the sample reflects both the
distribution of students between schools and the gender balance within each
school. For practical reasons, the final number of interviewees within each of
the eigth school/gender dividions did not always match the target number
exactly. however, deviations were small enough to be ignored for practical
purposes.
As a general rule of thumb, the statistical significance of
the results are only affected by absolute sample size, and not by its relation
to the size of the universe. However, this is no longer the case when the
sample exceeds about ten per cent of the universe. In such cases, the margin of
error is smaller than the absolute sample size would suggest. In the Student
Survey, the sampling percentage is as high as 42, which means that the margin
of error is the same as in a sample of about 450 drawn from a 'large' universe
such as the entire population of a country. The significance levels reported
for results of the Student Survey are in this sense underestimated.
The sampling process differed between schools, partly as a
result of the different willingness to excuse students from classes. In the
school on Santo Antão, ordinary classes finished two weeks earlier than
planned before the summer. This caused considerable difficulties and probably
led to biases in the sample. While interviewees were not selected on a truly
random basis in any of the schools, there is no reason to believe that there
are systematic biases in the sample in the other three schools. A truly random
sample means that every individual in the universe has an equal probability of
being sample, and that that this probability is independent of the sampling of
other inividuals.
Non-response was only a problem at the vocational school
(EICM), where a considerable share of the selected students did not show up for
their scheduled interview. At the other schools, interviews were not scheduled
beforehand, and very few students refused to be interviewed when they were
asked. |