Sample size
A small sample size might be appropriate in a study when there is a higher level of accuracy needed for the results to be valid, if the population of the research is small in size, and if the study is conducted through experiments.
A large sample size might be appropriate in a study when the size of the population under investigation is large, and if the survey is conducted through observation.
The benefits of experimental studies are that the researcher can remove or control factors outside of the independent and dependent variables that can affect the results of the studies and the ability of these experiments to be repeated several times to verify the cause-and-effect relationship. There are limits to an experimental study is that:
The more an environment and variable are controlled in the experiment, the more the results begin to no more show model real-world conditions.
Experimental studies are subject to human error. Researchers can make the conditions of the experiment conducive to the outcome or result they desire. Also, if the human subjects or participants in the research are aware of what is being studied, they can behave in manners that cause the data collection to become biased, where the result favors one outcome more than another.
The way the experimental and control groups are formed, especially when involving human participants, matters. These sample groups should reflect the general population that will ultimately be impacted by the results of the experimental study. Certain studies depending on the specific criteria and the potential risks surrounding these studies may have a limited selection of volunteering participants, resulting in a small sample size. This can also limit the interpretation of the data.
Whereas the benefits of observational studies are that it is useful when conducting experimental research would be deemed unethical. Limits of observational studies are:
Researcher bias can factor into the results hence compromising the results.
The data collected sometimes rely on the researcher’s interpretation.
The sample is not always representative of the population in that in some studies; the participants may be chosen simply by chance. For example, stopping people in a grocery store to taste two brands of orange juice to determine preference relies on chance.