Exam 4: Measurement Validity

A scale that is found to have high validity will inevitably also have high replicability.

During a lecture about designing your first research study, your professor (who is also your research advisor) told the class that in general, using a previously validated scale (when there is one available) is preferable to creating all of your own items. However, in a different lecture, your advisor mentioned that just because a scale is validated does not mean it will always be valid. Based on what you know about validity, which statement is correct? In what situations is the first statement better advice? In what situations is the second statement better advice?

Discuss the meaning of scale validation as a "continuous process."

When discussing the validity of a scale in terms of its relationship to a theoretically related outcome, _____(a)_____ refers to the type of validity in which scale and outcome are assessed simultaneously, while _____(b)_____ refers to the type of validity in which the outcome is assessed some time after the scale.

(a) Describe what is meant by heterotrait-monomethod, monotrait-heteromethod, and heterotrait-heteromethod in the multitrait-multimethod matrix approach. (b) Explain what these concepts supposed to assess.

Concurrent and predictive validity:

In a study of political preferences, Delilah assessed Democrats' and Republicans' attitudes toward the current US President (a Republican), as well as attitudes about various hypothetical governmental policies. This example reflects what type of construct validation?

People's responses on a scale can be influenced by systematic biases that are completely irrelevant to the construct that the scale was intended to measure. Specifically, imaging a researcher is asking for help in developing a scale assessing people's enjoyment of noxious or foul odors. Describe two threats to measurement validity that might influence responses to this scale that have nothing to do with people's penchant for stinky smells. In your response, be sure to (a) provide an example item for the scale that you will use as a reference, (b) identify and describe each validity threat, and (c) discuss how each validity threat would manifest in relation to your example item (i.e., explain how each threat would specifically bias responses, and in what direction).

Of all the various kinds of validity (predictive, convergent, face, content, discriminant, etc.) are there some forms of validity that are more important than others? Are they all equally important?