A study by Rachman in 1974 determined that, since fear and anxiety are closely related, subjects have a difficult time determining which is which. Thus, subjects in a fearful situation such as a radiology lab may have a tendency to over-estimate their levels of anxiety, due to their inability to distinguish fear from anxiety. On the other hand, smaller levels of anxiety may be undetectable to the subject, if the corresponding feeling of fear is absent (Ewert, 1986).
Additionally, survey research has been consistently been doubted due to the inability to prove a subjects response. Since surveys rely on the input of subjects, there are always slight possibilities that the subjects will lie, or be otherwise inaccurate in their responses. A study by Epstein in 1976 suggested that anxiety in humans is related to ideas of ego, self-esteem, and are associated with a weakness or inability to cope. Certain subjects who feel this way may inadvertently deny their own feelings of anxiety in a radiology department, because admitting the anxiety exists would threaten their self-concept (Ewert, 1986). In any of the above cases, the answers would be used in data collection, but would not be a true representation of anxiety levels.
A second type of research design is that of the experiment. In this design, the researcher manipulates an independent variable in