desired test results (Honts and Perry, 1992), and if this can be done intentionally, it might also be done unintentionally by an examiner who holds a strong expectancy about the examinee’s guilt or innocence (we discuss the expectancy phenomenon later in this chapter). Even if this calibration is not influenced by an examiner’s intended or unintended bias, it may be tipped one way or another by subtle variations in the ways an examiner introduces or conducts the test (Abrams, 1999). This source of inconsistency and potential unreliability in test administration was a stimulus for developing comparison question testing techniques that standardize the relevant and comparison questions across examinations and examiners. For example, directed-lie comparison question test formats have been advocated as superior to probable-lie variants because in the latter format, “it is difficult to standardize the wording and discussion of the questions” (Raskin and Honts, 2002:22). Concealed information test formats have also been advocated as superior to comparison question formats in this respect.

While orienting theory appears somewhat more plausible than the theories that underlie comparison question approaches, using the theory in devising polygraph procedures is not without problems. In particular, it is not clear how differences in stimulus familiarity affect orienting responses. Descriptions of this theory usually start with the assumption that responses to familiar and important stimuli will be different from those to novel, irrelevant stimuli, but in fact, the characteristics of stimuli should be thought of as a continuum rather than a dichotomy. That is, some stimuli are highly familiar and relevant and attract strong orienting responses, while others are moderately familiar and might or might not attract these responses. Orienting responses to familiar and important stimuli might generalize to other similar stimuli in ways that would make it difficult to distinguish true orienting responses from those bought on by stimulus generalization. For example, suppose a murder is committed using a nickel-plated revolver, and suppose an examinee owns an unregistered pistol (a blue-steel semi-automatic). That examinee might show enhanced responses to a variety of questions about handguns, even though he has no concealed information about the actual murder weapon.

Theoretical Limitations

The possibility of systematic individual differences or variability in physiological response has not been given much attention in polygraph theories. For example, the unresolved theoretical questions about the basis of inferences from the polygraph leave open the possibility, discussed below, that responses may be sensitive to effects of examiner expectations or witting or unwitting biases or to examinees’ beliefs about



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement