Generating random numbers in sample surveys
Abstract
A number of research activities taken in this 21st century are based on a selected part of the population and not the population as a whole. Automating the process of sample size determination linked with generating numbers random enough to fit the survey objectives is likely to offer great help to researchers of different background through reduced cost in terms of time and money.
This study took advantage of the existing technology to design a tool capable of determining sample size for finite populations given the desired level of precision. The tool was designed using C Language due to its rich set of Program Files, Functions,
Flow Controls, and the fact that by the time of generating the tool, it is the language the researcher was familiar with. Dev C++ also provided a conducive environment for generating the code.
It is capable of determining sample size and latter generate random numbers for three study designs i.e. Simple Random Sampling, Stratified Sampling and Cluster Sampling for the case of finite Populations. The generated numbers are then saved in a CSV file in the folder where the code is saved to provide room for the researcher to have access to the generated numbers after compiling/exiting the code.
The code/tool doesn’t cater for all designs that may be of interest to the researcher. For example Systematic sampling was not considered. It is recommended that the study is extended to cater of researchers who may have interest in such survey designs.