No. It measures physiological responses that can reflect anxiety or fear as well as deception, and scientific reviews found substantial error.
Most American courts exclude them or admit them only in limited circumstances, such as when both parties agree.
No. False negatives occur, so a passed polygraph should not be treated as proof of innocence in a cold case review.
Yes. Long, stressful sessions and being told one failed have preceded some confessions later shown to be false.
Generally not. Results do not prove guilt or innocence, and publishing them can unfairly damage a person's reputation.