A potential confound in field research on polygraphy. If cases are selected for research that use confession as a form of confirmation, then the study sample may be biased for the following reason. Standard practice in polygraphy is to only interrogate after a deceptive outcome on the examination. Therefore, confessions will only be obtained from examinees who failed the examination, but not from those who managed to defeat the examination. The sample will then represent those cases in which the examinee was caught by the original examiner, rather than all cases in which the examinee was deceptive. It has been asserted that the use of the confession-verified cases in blind scoring studies to assess polygraph validity may overestimate the accuracy of the polygraph, because they may have charts in which deception is the easiest to interpret. Most field studies that have examined this source of research error have not a meaningful effect, though the issue is still hotly debated. See: Horvath (1977); Honts (1996); Iacono (1991); Krapohl, Shull & Ryan (2002); Patrick & Iacono (1991); Raskin, Kircher, Honts, & Horowitz (1988).
