bias

What Is Bias in Polygraph Research?

Bias in research is a source of systematic error that can influence experimental outcomes in a consistent direction, producing results that do not accurately represent reality. Understanding and controlling for bias is essential to conducting and evaluating valid polygraph research.

Sources of Bias

Bias can enter polygraph research through numerous pathways. Sampling bias occurs when the research subjects are not representative of the population the study intends to generalise to — for example, using only college students to validate a technique intended for criminal investigations. Expectancy bias arises when researchers' or examiners' expectations influence how data are collected, scored, or interpreted. Selection bias can occur when using the confession criterion for case selection, as this may systematically exclude cases where the examiner made an incorrect diagnosis.

The Confession Criterion Problem

One of the most discussed sources of bias in polygraph field research is the confession criterion. If only cases confirmed by confession are included in a validation study, the sample may be biased toward cases where deception was relatively easy to detect — because confession typically follows only after a deceptive outcome was reached. Cases where deceptive examinees successfully defeated the examination would be systematically excluded. Most studies examining this specific concern have not found it to produce a meaningful effect, but the issue remains debated.

Controlling for Bias

Researchers attempt to control bias through careful experimental design: random assignment of subjects, blind chart analysis protocols, pre-specified (a priori) analytical methods, and appropriate control groups. The modern emphasis on evidence-based practice in the polygraph field reflects an ongoing commitment to minimising bias in the research that underlies professional standards.