2.5 Measuring and scaling
2.5.1 Sampling types
After defining how the data are going to be collected the next
consideration is how to select a sample of the population of interest that is
truly representative. In fact, it would be very costly and time-consuming to
collect data from the entire population of a market. The population will be
sampled by using a sampling frame. A sample is defined as «a subset of a
frame where elements are selected based on a randomized process with a known
probability of selection» (OECD Glossary of statistical terms, 2001).
There are various types of sampling frame.
At the beginning, for the present study, the researcher thought
of two different types of sampling: quota sampling and convenience sampling.
Both of these sampling are non probability; Castillo, Joan
Joseph (2009) defines non-probability sampling as «a sampling technique
where the samples are collected in a way that does not give all the individuals
the same chance of being selected». As a consequence it is true to assume
that probability sampling is often more representative; however it is also more
complicated to set up. Indeed, the researcher was facing many limitations like
time, workforce and money; as a result it is almost impossible to randomly
sample the whole population. For the present study, the researcher decided to
use a non probability sample based on the accessibility of the samples. As it
was said before, the researcher thought of two different sampling frames: quota
and convenience sampling.
Quota sampling permitted to obtain representative data of the
overall population by divided it by the most important variables. This is quick
and easy to set up. However, this type of sampling as it is not made randomly;
the risk of bias is rising. Unlike quota sampling, convenience sampling permits
to gather quickly a large amount of information and it is readily available.
However, the risk of no response is important and the researcher has to make
sure that all the respondents have equal chance to be interviewed.
For this study, convenience sampling was finally chosen. In
fact, convenience sampling would allow an easier and quicker establishment;
whereas quota is more time consuming as the researcher will need more knowledge
about the population for the stratification.
2.5.2 Sampling size
After defining the sampling type, it is important to decide of
the sampling size. The most obvious reason is that the biggest the sample is
the better is. In fact, larger sample tend to be more similar to the population
and, as a consequence, permits to get more representative information.
However, large sample are costly and more time-consuming than
smaller sample. As a convenience sampling was used for this study, we need a
relatively large sample according to time and a money constraint, the research
has tried to get the larger sample as possible.
2.6 Data processing and analysis
The questionnaire was an online one; therefore it was easier
for the researcher as the distribution of the invitation is very rapid (email
with hyperlink), the data could be downloaded and imported to SPSS and it is a
low cost method of gathering information.
After gathering all the completed questionnaires from the
respondents, the total responses for each item were obtained and tabulated.
This would allow the researcher to get various data, which would be analyzed in
order to validate or not the hypothesis.
In order to analyse the relation between the independent and
the dependent variables, the researcher tabulated the data in SPSS software in
order to do a linear regression analysis.
Regression is a statistical test designed to predict a
dependent variable from one or more independent variables, this would allow the
researcher to test the different hypotheses according to the different
variables. (Alan O. Sykes 1986) Each hypothesis was test with linear regression
analysis and the obtained results were described and explained in results
part.
| 
