pailryder wrote on Feb 1^st, 2017 at 1:53pm:

On more than one occasion I was contacted after several years and thanked by couples who felt fidelity testing improved or saved their marriages.

I am not at all surprised to hear this.

Since the polygraph gives a random result, of course some subjects are going to be deemed truthful. Assuming that some subjects of polygraph fidelity testing are in fact faithful and truthful, while others are adulterers hoping to skate by, this yields four cases: Tt, truthful tested truthful; Tf, truthful tested false; Ft, false tested truthful; and Ff, false tested false. The value of the testing can be demonstrated ONLY by the lack of Tf and Ft cases.

Testimonials from couples who felt fidelity testing improved or saved their marriages demonstrates only the existence of a non-zero population in case Tt. Unfortunately the world is full of innumerate people who don't understand, and gullible people who don't care, what any of this means.

I would be surprised to hear that every person who claimed that fidelity testing improved or saved his marriage was determined, through independent means, to have been telling the truth. In fact, I would be astonished at any independent collaborating evidence for any lie detector test at all.