Sampling Strategy and Sample Size for a Quantitative Research Plan

Population

Notably, the population of a given research is constituted by all possible individuals who could be subjects of the research. The target population for the forthcoming study will comprise of all the salaried persons in a given locality defined by high crime incidence. Specifically, the population will comprise of all those have resided in the locality for at least six months and have been getting salaries for at least 12 months prior to being interviewed by the researcher.

Population Size

There is no census that has been conducted in the locality in the recent past. That means that the precise size of the population at present cannot be established. Even then, it is estimated that in the locality, there are 3,500 who will have resided in the locality for at least six months and been getting salaries for at least 12 months prior to being interviewed by the researcher.

Sampling

Samples are subsets of the proposed populations for particular researches. The subsets are deemed to be representing their larger populations according to Bartlett, Kotrlik and Higgins (2001). Notably, the data gathered from the subsets is referred to as statistics, which are used in making specific inferences regarding the populations represented by the subsets. Sampling is the processes of selecting the subsets from given research populations. In the forthcoming research, the sampling will be probability based owing to various reasons.

First, the target population is already well-defined and known. Second, probability sampling will be done since the researcher has a detailed sample-frame, or list, of the population already. Third, the research will entail statistical analysis and only probability sampling methods are well-suited for the analysis. Lastly, the methods are less susceptible to bias than non-probability sampling approaches.

Sampling Type

The forthcoming research will entail the usage of a random sampling approach, simple random sampling (SRS). Notably, the specific sampling techniques adopted by researchers are largely dependent on the forms of interviews to be executed according to Bartlett, Kotrlik and Higgins (2001).  In the forthcoming research, the researcher will carry out structured interviews. SRS is well-suited for such interviews.

As noted before, there are other reasons why SRS will be suitable for the research. First, the target population is already well-defined and known. Second, the researcher has a detailed sample-frame, or list, of the population already (Emmel, 2013). Third, the research will entail statistical analysis and SRS is well-suited for the analysis. Fourth, SRS is less susceptible to bias than non-probability sampling approaches. In the research, the SRS will be executed by computer-based programs. Fifth, the sample will be assembled easily and fairly. Sixth, SRS will be highly representative of the research’s target population. Owing to SRS’ representativeness, it allows for the making of generalizations from sample results to the target populations (Brewerton & Millward, 2001).

How the Sample Will Be Drawn

Given that SRS will be employed in the upcoming research, every member of the research’s target population will stand the same likelihood of being chosen to be one of the research’s subjects. The sampling will be executed in a lone step, will each of the possible subjects chosen independently of all the other possible subjects (Brewerton & Millward, 2001).

Specifically, data on the possible subjects, those who will have resided in the locality for at least six months and been getting salaries for at least 12 months prior to being interviewed by the researcher, will be obtained from the labor office in the locality. The names of the possible subjects and their telephone contacts will be extracted from the labor office’s data base. The names and the corresponding telephone contacts will be saved in a computer. The computer will be used in aiding the random choosing of the forthcoming research’s sample.

Sample size

When executing studies, researchers are keen on the number of responses that they really require according to Bartlett, Kotrlik and Higgins (2001). The following formula has been used in computing the required size of the upcoming research’s sample.

Required Sample Size = StdDev * (Z-score)²  * (1-StdDev)  /  (error margin)²

The error margin, or confidence interval, that will be allowed in the research will be ±5%. The standard of deviation (StdDev) expected in the responses that will be given by the selected subjects will be 0.5. Notably, the 0.5 StdDev will make certain that the sample will be sufficiently large. The expected confidence level in the research will be 95%. The Z-score relating to that confidence level is 1.96. Thus, the required sample size in the research will be computed as:

Required Sample Size = 0.5 * (1.96)²  * (1-0.5)  /  (0.05)²

=  384.16

=  385 respondents

In the light of the research’s 3,500-person target population, the 385-respondent sample will be proper. The sample will ensure that the allowed error margin, or confidence interval, will be minimal. The sample will allow the researcher to be highly confident that the real mean will fall within the interval (Brewerton & Millward, 2001). As well, the sample size will ensure that only a highly limited variance will be expected in the responses. Overall, the sample size will yield markedly precise results as demonstrated by Bartlett, Kotrlik and Higgins (2001).

Scroll to Top