3.1.3 Target population and
sampling
The population of this study was made up of Entrepreneurs or
school owners, class teachers and pupils. The large number of private schools
in Kibera catering for various social classes prohibited a systematic sample.
In fact random sampling was applied in order select our sample schools and the
respondents. This was conducted in such a way that each has an equal and
independent chance of being selected (Best, J &Kahn 2003:13) Participants
were asked to volunteer for the study and a simple random sample was done.
We were very much interested in private school
entrepreneurship offering services to the poor. As such, twenty school owners
were chosen. Apart from this number, five teachers each and five pupils each
were selected from five different schools. Thus, the total population of 70 was
administered interviews and questionnaires.
3.1.4 Reliability and Validity
of the research
Reliability and validity are considered highly essential in
social science research. They contribute to reassure that the instruments used
in carrying the research were appropriate and effective with close regards to
the central question of the study. We therefore chose to use the above
mentioned instruments in the process of data collection in Kibera with close
regards to these two parameters. In another words, it's advised that a quality
research should produce similar results under constant conditions on all
occasions. Bell (2005:117)
An utmost precaution was taken before the beginning of the
study, which was to see that our instruments fit perfectly well with the
research. We took upon the responsibility to ensure that our instruments of
data collection for entrepreneurs, teachers and pupils were of as high quality
as possible both in terms of design and content, and as unobtrusive and
inoffensive as possible. (Fogelman and Comber, 2007:129)
Another nonetheless major concern was equally that related to
the validity of our research with primary education and entrepreneurship in
Kibera. The more valid a research is, the more credit it gives to the entire
process of measuring various aspects of it content. A critical assessment and
reassessment of the questionnaires and Interviews were done by colleagues and
friends of the EG West Centre under the supervision of Dr Dixon, to help
determine that all the questions measure what they were supposed to measure.
Bell (2005:117)
However, being more complex on its own, measuring the validity
evidence of our research in Kibera required us stress much of certain aspects
of the instruments that were used. This was done with close respect to
emphasise that is usually laid on this issue by scholars when they argue that
validity evidence is based on three broad sources namely: content, relation to
other variables, and construct (Best, J &Kahn 2003:282-286)
While not analysing these three latter parameters of validity
as totally separate entities, we however designed our research in such a manner
to arrive at credible conclusions. (Sapsford and Jupp 1996, cited in Bell 2005:
117-118)
In the perspective of rendering the research more accurate and
realistic, the triangulation of multiple variables of the respondents' answers
were equally done, thus making the overall study reliable and valid.
| 
