Final answer:
The best control for sampling bias is random selection from the population, as it gives all members of the population an equal chance of being included in the study sample, thus minimizing the possibility of bias. Random assignment helps to control for confounding variables within the experiment but does not address sampling bias directly.
Step-by-step explanation:
The best control for sampling bias is b) random selection from the population. This is because random selection ensures that every member of the population has an equal chance of being included in the sample, which is crucial to avoid sampling bias. Random selection helps to create a sample that is representative of the entire population, thereby making the study's findings more generalizable.
Choice a) random assignment is also vital in the context of an experiment as it reduces the risk of confounding variables by ensuring that all groups in the study are comparable. Random assignment involves assigning the selected participants to different groups, such as treatment and control groups, in a random manner.
Choice c) double-blind study, while essential for eliminating both subject and experimenter bias, does not primarily address sampling bias but rather controls biases during the experiment phase.
Choice d) Hawthorne effect, is not a control for sampling bias; it refers to a change in behavior by study participants due to their awareness of being observed.