Associations between Quantitative and Qualitative Job Insecurity.

BUACC5931 2019 Semester 1, Assignment 1
Associations between Quantitative and Qualitative Job Insecurity and Well-being
Q1: Sample size
The sample size for this study is fifteen thousand employees selected from a total of 69,000 bank
employees (about 21% of the employees). Is a sample of this size necessary? Give your reasons.
A small sample size is not ideal for a study has it has large variability. A sample size is believed
to be small if it is not representative of the entire population. A small sample size also has a
lower power. The ideal sample size should be more than 30 or more than 10% of the entire
population size.
The above sample size of 21% is too large for the research. This will lead to overspending hence
making the research itself expensive. The researchers will also have to use much time collecting
and analysing data. The large sample size also leads to overrepresentation of the target
population. It is therefore not necessary to have such a huge sample size for data collection
(Francis, 2010). With the large sample taken, it becomes difficult managing the data collection
materials as one cannot follow up accurately with individuals under study in some situations
resulting to loss and spoiling of data collection tools. From the large number of sample, it
becomes difficult to control the sample respondents. It may also be difficult for the general graph
illustration as it is too large (De Witte, et.al., 2010).
However, when a more accurate data is required with the availability of enough resources
allocated, adequate time given for the research, highly skilled personnel available that can help
handle the data collected then a large sample size can be convenient for use. This will increase
Page 1 of 10

BUACC5931 2019 Semester 1, Assignment 1
the total number of employee representation reducing the sample error that may result from a
small sample taken.
Q2: Sampling method
What is the current method of sampling? What are the advantages and disadvantages of the
current sampling method?
The current method of sampling is cluster. This is where the sample is randomly picked from the
different groups spread over the area/country under study. For instance, picking samples from the
63 different banks across Belgium. The sample size of 21% is equally taken to represent the
entire population (Bland, 1998)
The cluster sampling method has the following advantages;
i. Allows for the usage of less amount of funds in conducting the research. The small
samples are easily managed in as fewer resources are used as compared to systematic
sampling technique;
ii. Cluster sampling reduces variability as it provides estimation sample that’s more
accurate;
iii. Its more feasible as it represents the whole population for instance, all the 63 banks of
Belgium are equally represented reducing biasness;
iv. It covers wide area as all the 63 banks across the country are taken. This reduces biasness
that could come from conducting a research only in one area. It also ensures that different
views are collected over a wider area leading to equality and proper distribution and
Page 2 of 10

BUACC5931 2019 Semester 1, Assignment 1
v. And creates large data sample. This improves the accuracy and reliability of the data
collected. The large data given will also open scope to wider views from the respondents.
The sampling method however has the following disadvantage:
i. With cluster sampling, it is easier to create biased sample. In a case where the sampled
population has a biased opinion, then the entire population becomes of the same opinion;
when the 21% sample majority has the opinion of poor working conditions then the whole
population of workers will have by default be involved;
ii. Prone to sampling errors as it is majorly based on the researcher’s skills and how familiar
he/she is to the area under study;
iii. It can work well when individuals/ people are grouped by units, without which it cannot
work;
iv. Lowest number of cases required for accuracy and to avoid overlapping. Without this,
there is a chance of obtaining one sided data;
v. The data obtained will only be specific to the population group. For instance, the factors
affecting the future of the employees in a banking industry cannot be applied in other
sectors of the economy or other countries apart from the bankers in Belgium;
vi. Cluster sampling must have sampling equality to succeed. For instance, taking 21% from
some banks and 5% from other banks will lead to underrepresentation of some areas and
vii. Each and every cluster may have an overlap that reduces the integrity of the data
collected. (Bertrand, 2004).
Page 3 of 10