Reaction Time

Rationale

Why should RT diagnose deception? Everything else being equal, the more complex a mental process is, the longer it takes to unfold, In general, lying is a more complex process than truth-telling. Hence, lying takes longer to unfold. It follows that RT can empirically distinguish lying from truth-telling.

Why is lying a more complex process? There are several reasons. For one thing, lying involves extra cognitive operations. In an honest respondent, an inquiry automatically activates knowledge, which is then directly communicated. In a dishonest respondent, however, knowledge activation is followed by deliberation over whether to tell a lie, and once the decision is made, by an attempt to construct one, entailing the suppression of activated knowledge and its replacement by an appropriate fiction. Moreover, the ongoing challenges of keeping a false narrative straight, of monitoring and managing one's demeanor, and of dealing with the fear of exposure all compound the complexity of lying.

The greater complexity of lying makes it more difficult. Hence, for lying to succeed, relatively more mental effort must be exerted and relatively more mental resources must be available. Tellingly, slower RTs are often accompanied by increased errors. Note that, given the speed-accuracy trade-off, attempts to reduce errors will retard RTs, while attempts to hasten RTs will increase errors.

Measurement, Magnification, Manufacture, and Modification Detection Techniques

RT-based lie detection techniques can be grouped into four classes. Those in the first class involve measurement. They simply capitalize on lying being naturally more complex than truth-telling. Those in the second involve magnification. They amplify the foregoing natural difference. Those in the third involve manufacture. They engineer an entirely artificial situation in which slower RTs diagnose deception. Those in the fourth involve modification. They take an existing paradigm and add RT measurements to it.

On occasion, techniques in all classes detect deception comparatively well. The jury is still out, however, on whether they will ultimately prove robust to countermeasures or practical to implement. The techniques are described as follows:

Measurement. When respondents lie, their verbal replies to spontaneous inquiries take longer. Yet this effect seems confined to inquiries that arouse emotion; plain facts can be lied about just as quickly. However, if verbal inquiries are repeated and verbal responses to them dichotomized (for example, as yes/no), or if statements are presented on a computer for digital classification, then the effect emerges solidly for plain facts too. True, prior practice at lying, by decreasing RTs generally, attenuates diagnostic accuracy. However, taking errors into account alongside RT augments accuracy.

...