Research sampling in the social sciences: random and non-random methods
The selection of the sample determines who the results of the study really relate to.The number of surveys alone is not enough: it is necessary to indicate the population, how to reach the participants, the criteria for participation and the limitations of the procedure adopted.The guide below helps to organize these decisions without pretending to be representative, which the study does not provide.
Key information
- ✓First define the population, then select the sample.
- ✓Don't call a trial representative without a methodological basis.
- ✓Establish both the selection method and the criteria for participation.
- ✓Describe the limitations of the test directly in the methodology and conclusion.
Population, sample and unit of study
Populations are the collective that you want to draw conclusions about, but the test involves people or objects actually participating in the study. You should also name the unit of analysis: it can be a student, employee, company, document, or event.
Random selection and its conditions
In random selection, each individual has a known likelihood of being in a sample, which requires a lottery operation, or as complete a population list as possible. Simple, systematic, or layered lottery can limit bias, but only if the procedure has actually been performed and described.
Non-random methods: targeted, quantitative and snowball
Targeted selection works when you're looking for people with specific experience. Quantitative selection reflects selected populations, and snowballing makes it easier to reach hard-to-reach groups. These methods are useful, but they don't automatically generate results for the entire population.
How to Describe a Job Exam
Please indicate the place and time of the study, the method of recruitment, the criteria for inclusion and exclusion, the number of people invited and the number of complete responses, then characterize the study with the characteristics relevant to the problem and explain the limitations of the method of selection.
Summary
A well-described trial does not have to be perfect, but it has to be logically related to the purpose of the study. The reader should know who might have been in the study, how the participants were selected, and how this decision limits the interpretation of the results.
Frequently asked questions
How many respondents do you need for thesis?
There's no single number for each job, but it depends on the goal, the population, the planned analysis, and the university's requirements, and it should be agreed upon before data collection starts.
Need support with your work?
Describe the topic, scope and deadline. We will review the materials, recommend the right support and prepare a clear quote.
