Validity and reliability are both important terms for experimental design and are both markers of a good study. A study is valid if it measures what it is supposed to measure and reliable if it produces similar results every time it is conducted. If it is not valid, then it cannot answer the hypothesis or be applied to the real world; furthermore, if it is not reliable, then the results cannot be trusted.

A study is valid if it measures what it is supposed to measure and reliable if it produces similar results every time. The choice that stated “a scale that shows you that you are 50 pounds every time you step on it (i.e. reliable), but you know that you are really 150 pounds (i.e. not valid)” is reliable but not valid.

A study is valid if it measures what it is supposed to measure and reliable if it produces similar results every time. The choice that stated “a scale that shows you between 140-160 pounds every time you step on it (not reliable) and you know you weigh 150 pounds (valid)” is valid but not very reliable.

Dr. Andrews’ study lacks external validity because his findings cannot be applied to the greater population. There is something about his sample—in this case, the fact that they all came from the United States and all were raised to speak English—that differs significantly from the greater population (i.e. not everyone in the world is a an English-speaking, American citizen); therefore, the results cannot be generalized to the entire population. A problem with internal validity would involve a problem with the proposed relationship between the measured variables (e.g. if Dr. Andrews had found that age and phoneme usage were completely unrelated or if Dr. Andrews discovered his methods for measuring phoneme acquisition were inaccurate). Anonymity is not mentioned in the description, but we can assume Dr. Andrews did not reveal any personally identifying information about his subjects; furthermore, anonymity would be unrelated to whether or not his results can be generalized to a larger group.

A test is reliable if it consistently measures what it is intended to measure. Reliability can be tested by giving the same test to the same group of people at two different points of time or giving two forms of a test to the same group of people at different points of time. Reliability is associated with the words "consistency" and "stability." All of the other choices are definitions of validity. A test has construct validity if it measures what it says that it measures. A test has content validity if it measures all aspects of a domain. A test has criterion validity if it can make a prediction based on test scores. 

The purpose of random sampling methods is to ensure that one individual does not have a greater chance of selection than any other. Although random sampling reduces the effects of random variables, like error, they are not completely eliminated. Representation is not guaranteed, but is likely heightened. Although randomness does not ensure conclusive results, it helps to validate any results of a study against one that is not random.