Is it possible to have a valid test that is not reliable? Why or why not? what are some general differences between external and internal validity? how can internal validity be improved? Differentiate between probability and nonprobability sampling plans. Review briefly their general functions and limitations.