In academic research, the solution that a study aims to provide, is usually tailored towards a particular group of people. “Particular” because it is expected of a quality research to be as specific as possible.
Take a look at these five (5) examples:
In understanding the academic decline among students whether at the college or university level, a researcher can only understand this association by extracting information from the affected party who are the student, Right?
In a university where there is an increase in the prevalence of sexual violation among students, there could be a need to understand the association between the sexual violation and the academic performance of the affected. The group likely to provide a valid insight into the association are more likely those students who have experienced sexual assault in any of its forms in the walls of the learning institution.
Examine a situation whereby self-harm victims taken care of by newly qualified mental-health nurses show positive responses to certain care strategies. In a bid to know which of the care strategies help with self-harm episodes, the best group to provide this responses are either the victims themselves or the newly qualified mental-health nurses administering the care.
In a modern integrated-care settings, the understanding of the roles as well as the opportunities and challenges faced by social workers can best be provided by social workers alone. Lastly, examining the impact of the quality of life of police officers on their sexual satisfaction requires interviewing police officers as this group is the only group that can provide valid information.
Before proceeding with this article, here is a list of services offered by Craig Educational Consult in order to improve your academic performance.
All five (5) examples above, reveals the respective population (broadly), samples (specifically) that the studies focus on. It justified the need that studies are focus-oriented and meant to solve the problems of specific groups. This article aims to explain the process through which researchers arrive at the ideal sample size for their studies.
Its content will include the following:
What is a Sample?
What is a Sample Size?
How do you calculate Sample Size?
WHAT IS A SAMPLE?
It is impossible for researchers to interview large populations due to the effort, feasibility, money and time involved. It is thus a usual practice for researcher to work with samples of the identified population and use it to make estimate about the entire population. The latter can be conducted quickly and data managed easily.
By sample, this article means a controlled representation derived from a larger population. The samples are extracted from the main population and possesses the trait of the parent population. Simply refer to them as subsets of a larger population.
WHAT IS A SAMPLE SIZE?
In order for any sample to statistically represent a population, it is important to use the appropriate sample size. Sample Size is defined as the number of observations (samples) selected from a population for a study which directly influences the accuracy, reliability and generalizability of the research findings.
HOW DO YOU CALCULATE A SAMPLE SIZE?
Five (5) expectations are required of a researcher whenever the need to calculate sample size arise. They include the need to provide information regarding the statistical analysis to be applied; determine the acceptable precision level; decide on the study power; specify the confident level; and determine the magnitude of the practical significance difference.
On statistical analysis, calculating sample sizes involves adopting formulas as long as the variables in the formulas are provided in the study specifications. Cochran formula; Taro Yamane; and Steven Thompson formulas are a few frequently adopted by researchers.
The acceptable precision level address the margin of error since errors are unavoidable in research. It is a percentage which show the statistical inference about the confidence that the number of respondents accurately represents the opinions of the whole population.
Specifying confidence level is important in sample size calculation because it shows the extent to which the sample represents the total population within the selected margin of error. The most frequent confidence level is 95 per cent. 90 and 99 per cent levels have also been reported in several study.
Using the Cochran formulas as an example
no = Z2PQ / e2
where no = Sample Size;
Z = Z value at a given confidence level;
P = Estimated Proportion of an attribute that is present in a population;
Q = 1 – P
e = Precision level
In the formula above, there is a parameter Z score. The Z score is usually obtained from the confidence level. The table below presents the Z score of the most used confidence level
| Confidence level | Z-Score |
| 80 per cent | 1.28 |
| 85 per cent | 1.44 |
| 90 per cent | 1.65 |
| 95 per cent | 1.96 |
| 99 per cent | 2.58 |
As an example, a study is interested in examining the knowledge and usage patterns of family planning methods among married women in an area (*Bayreuth). The confidence level is 95 per cent, margin for error is 5 per cent, and the estimated proportion of married women in the population is 40 per cent.
Using the Cochran formula above, finding the sample size is easy by substituting the values for the necessary parameters.
no = Z2PQ / e2
no = 1.96^2 *(0.40*(1-0.40)/0.05^2)
no = 3.8416 (0.40 * (0.60) / 0.0025)
no = 3.8416 (0.24/0.0025)
no = 3.8416 (96)
no = 368.79
no = 369
* Hypothetical
As a university student, if you constantly struggle with any aspect of your academic work (course works, assessments, tests, essays, assignments, research, analysis and writings), you can hire a private tutor at Craig Educational Consult to help improve your performance. If you know any student in need of these resources, direct them here
