correlation
What Is Correlation?
Correlation is a statistical measure of the relationship between two variables — the extent to which changes in one variable correspond to changes in another. Correlations range from -1.0 (perfect inverse relationship) through 0 (no relationship) to +1.0 (perfect direct relationship). In polygraph research, correlational analyses are used to evaluate relationships between physiological measures, to assess interrater reliability, and to validate scoring methods.
Applications in PDD Research
Common correlational analyses in polygraph research include: measuring the Pearson correlation between two examiners' scores on the same charts (high correlation indicates good reliability), correlating physiological response magnitudes with deception status, evaluating the relationship between computer-generated scores and examiner-generated scores, and assessing test-retest reliability across repeated examinations.
Correlation vs. Causation
An important principle in interpreting polygraph research is that correlation does not establish causation. A strong correlation between a physiological measure and deception status does not prove that deception causes the physiological change — it only establishes that the two consistently co-occur. Establishing causation requires controlled experimental designs that rule out alternative explanations, which is why laboratory studies with random assignment complement field research in building the evidence base for polygraph validity.
Types Used in PDD Research
The most common correlation methods in polygraph research are the Pearson product-moment correlation (for continuous, normally distributed data) and the Spearman rank correlation (for ordinal or non-normally distributed data). Each is appropriate for different types of data and research questions.
Cross-references
More scoring terms in this glossaryNeed a lie detector test?
Book with an examiner trained to APA standards at one of our 140+ locations nationwide.

