The Research Design
The following should be considered in the design of research:
- Research details
- Statement of the problem
- Type of research
- Purpose of the study
2. Sampling and data collection methods:
- Statistical population
- Sample size
- data collection method
3. Data analysis method
Data Collection and Sampling Methods
Reasons for sampling:
– Unavailability of the whole society
– Reducing the duration of the study
– Increasing the productivity of human resources involved in the research
A collection of people, objects, etc. that have at least one common attribute.
A sample is a set of subjects that are selected from a larger group, or community so that it represents the quality and characteristics of that larger group or community.
It is a way of choosing a part of a population so that all possible examples have the same chance of being selected.
Types of Sampling
1. Simple RandomSampling:
- The researcher gives all the subjects in the statistical population the same chance of being selected.
- The bias is at its lowest level.
- However, it’s hard and we may not always be able to come up with desirable representatives of the population.
2. Systematic Sampling:
- Extracting members of the community with a particular pattern.
- It is usually good for market research and investigation of the customer perceptions
- It must be observed for probable biases
3. Stratified Sampling:
It is used when we have several different levels of subjects, such as seniors, middle and lower-ranked managers. 4. Single-stage and multi-stage cluster sampling (MCS multistage Cluster Sampling)
4. Multistage Cluster Sampling
It is used when we have heterogeneous groups. For example, investigation of the average monthly deposits of banks.
In general, “the largest sample size is appropriate,” but financial, human, and time limitations will always pose major challenges in this regard. Its influential factors include:
- Research method (for example, in causal research, a small number will be enough)
- Research goals (a researcher who intends to generalize the results of her research to the statistical community should choose a larger sample size).
- Financial limitations.
- The size of the statistical population (the lower it is, the larger the sample size should be)
- Percentage of error in results (usually in research, 5% of statistical error is acceptable)
- The effect of the independent variable on the dependent (the less the effect of the independent variable on the dependent one, the larger the sample size should be)
Validity is the creation of conditions that show the effect of the independent variable on the dependent one without the influence of the unwanted variables.
Is the result of the research effected by the influence of the independent variable on the depend on one?
To what extent can we rely on these findings?
To what extent can the research findings be generalized to a larger community?
Several factors affect the validity and need to be considered:
- Research period (events that occur during the research and affect the independent variable)
- Pre-test (for example, participating in practice entrance exams in research on the effect of entrance exams on behavioral abnormalities)
- Lack of accuracy of the measuring instruments (recording observations while being tired or using tests and methods that are not in line with the culture of the society)
- Sample attrition (overtime, subjects may change. For example, some of them may leave the test, especially in long-term research)
- Unbalanced groups
- Artificial test conditions
- The effect of independent variables on each other (for example, comparison of the effect of the required prerequisites in two training courses)