Chapter 31 — Self-Check Quiz

24 questions: multiple choice and short answer. Try them closed-book. The answer key is in the collapsed block at the bottom.

Multiple choice

1. The chapter's central claim about cognitive bias in forensics is that the threat comes primarily from: - A. Dishonest analysts who fake results - B. Lazy or undertrained analysts - C. Normal human cognition operating on ambiguous evidence under the influence of expectation, even in competent, sincere analysts - D. Broken laboratory equipment

2. "Systematic" bias differs from "random" error in that: - A. It is smaller in magnitude - B. It pushes judgments in a consistent direction, so making more measurements with the same expectant mind does not cancel it out - C. It only affects beginners - D. It can be detected by recalibrating an instrument

3. Confirmation bias, at the bench, primarily causes an examiner to: - A. Weigh evidence agreeing with their tentative conclusion as significant while rationalizing or discounting discrepancies - B. Refuse to reach any conclusion - C. Always exclude the suspect - D. Ignore the reference sample entirely

4. Contextual bias is best defined as distortion of a judgment by: - A. The quality of the laboratory's microscopes - B. Domain-irrelevant information from outside the evidence (case facts, suspect's identity, a confession, the desired outcome) - C. The examiner's eyesight - D. The age of the evidence

5. Which of the following is domain-irrelevant information for a latent-print comparison? - A. The ridge detail of the latent print - B. The substrate the print was developed on - C. The fact that the suspect already confessed - D. The clarity of the exemplar

6. The relationship between contextual and confirmation bias is best described as: - A. They are identical - B. Contextual bias supplies the expectation (trigger); confirmation bias is the mechanism that steers the analysis toward it (engine) - C. Confirmation bias only occurs without any context - D. They never occur together

7. The bias cascade refers to: - A. A waterfall of evidence at a crime scene - B. An expectation or irrelevant context introduced at one point spreading through later analyses, verifications, and re-examinations, so apparently independent agreements are one biased judgment amplified - C. The order in which insects colonize a body - D. A type of DNA contamination

8. When four examiners all conclude "match," this is strong evidence of accuracy only if: - A. They are all FBI examiners - B. Each reached their conclusion without knowing the others' conclusions and without the contaminating context that started the cascade - C. They worked in the same room - D. The first examiner was very experienced

9. In the Mayfield case, the verifications failed as a check because: - A. The verifiers used the wrong equipment - B. The verifying examiners knew an identification had already been declared (the verification was not blind) - C. There were no verifiers - D. The latent was a perfect, complete print

10. Domain-irrelevant context in the Mayfield case included: - A. The ridge detail of the latent - B. Mayfield's religion and his prior legal work for a convicted person — facts with no bearing on whether two ridge patterns correspond - C. The development chemistry used on the print - D. The Spanish police's dissent

11. The Dror experiments demonstrated, under controlled conditions, that: - A. Fingerprint examiners are usually wrong - B. The conclusions of qualified experts can be changed by domain-irrelevant context on the very same physical evidence - C. Fingerprints are not unique - D. AFIS is unreliable

12. Which statement about the Dror findings is the chapter's honest framing? - A. They prove the fingerprint method is junk science - B. They depend on one dramatic statistic - C. They show a portion (not all) of expert conclusions changed under context, establishing that context can change expert judgments — without implying most comparisons are wrong - D. They show examiners are incompetent

13. Context management is: - A. Telling the examiner everything about the case so they are fully informed - B. Deliberately controlling what information reaches the examiner — supplying domain-relevant data, shielding domain-irrelevant context - C. Managing the laboratory's budget - D. A type of chain-of-custody form

14. Sequential unmasking requires that the examiner: - A. Sees the reference and the case theory first, then the evidence - B. Analyzes and documents the evidence sample alone before any reference or case context, then has information revealed in a controlled order (most-biasing last) - C. Works only at night - D. Never sees the reference sample at all

15. Requiring an examiner to document their interpretation of the evidence sample in writing before seeing the reference is powerful because: - A. It is required by law everywhere - B. A time-stamped Stage-1 record cannot be silently revised to fit the exemplar later, making any shift visible and auditable - C. It speeds up the analysis - D. It eliminates the need for a reference sample

16. Context management matters most for which kind of task? - A. Reading a number off a calibrated instrument - B. Subjective comparison of ambiguous patterns (fingerprints, toolmarks, bloodstain interpretation) - C. Weighing a powder - D. Recording a temperature

17. A DNA mixture interpretation needs context management more than a single-source profile because: - A. Mixtures use different chemicals - B. Mixture interpretation involves judgment (number of contributors, which alleles to call) that can drift toward an expectation when the analyst knows the suspect's profile - C. Single-source profiles are never used in court - D. Mixtures are always inadmissible

18. The main reason most labs still have not adopted context management is best summarized as: - A. It does not work - B. A combination of the institutional bias blind spot, structural/resource constraints, and the absence of a forcing mechanism - C. It is illegal - D. It is too expensive for any lab on Earth

19. A "verification" that is not blind is, in the chapter's view: - A. A complete solution to bias - B. Part of the cascade, not a cure — it manufactures the appearance of corroboration while adding none of the substance - C. Always better than nothing - D. Required by Daubert

20. A foundationally valid method (e.g., latent prints) in a contaminated workflow can: - A. Never produce an error - B. Produce a confident false positive, because validity of the method does not protect the human judgment from bias - C. Only produce exclusions - D. Automatically correct for bias

Short answer

21. In two sentences, explain why an examiner's sincerity and belief in their own objectivity is not a safeguard against bias — and name the specific phenomenon that makes self-confidence in objectivity dangerous.

22. State the single most important consequence of the bias cascade for interpreting "several examiners agreed," and give the one condition under which their agreement becomes genuine corroboration.

23. Name the three stages of sequential unmasking in order, and state what most-biasing information should be withheld from the bench entirely.

24. Explain how the chapter's Case File re-reads the original "accidental fire" frame in the cold case — what role did that frame play, and what is the honest status it produces? (Do not claim the finding excludes or includes anyone.)

Answer key (click to expand) **Multiple choice:** 1-C · 2-B · 3-A · 4-B · 5-C · 6-B · 7-B · 8-B · 9-B · 10-B · 11-B · 12-C · 13-B · 14-B · 15-B · 16-B · 17-B · 18-B · 19-B · 20-B **Short answer (model points):** **21.** Perception and interpretation are a single process that reaches for context to resolve ambiguity, so by the time an examiner is "ignoring" the confession it has already shaped what they saw; sincerity does not interrupt that. The dangerous phenomenon is the **bias blind spot** — people concede that others are biased while believing themselves immune — which leads the most confident examiners to reject the very safeguards that would protect them. **22.** The consequence: **unanimity is not corroboration / agreement is not independence** — multiple examiners who each knew the prior conclusion (or shared the contaminating context) are one biased judgment echoed, not several independent ones. It becomes genuine corroboration only if each examiner reached their conclusion *blind* to the others' results and free of the domain-irrelevant context that started the cascade. **23.** **Stage 1:** analyze and document the evidence sample alone (no reference, no suspect, no theory). **Stage 2:** reveal the reference/exemplar, compare, and record the comparison conclusion before any case context. **Stage 3:** reveal limited task-relevant context only if genuinely needed (e.g., substrate). The **most-biasing information** — the confession, the suspect's record, the case theory, other examiners' results — should be withheld from the bench entirely (handled by a case manager). **24.** The "accidental fire" frame was not neutral; it was an **expectation recorded before any analysis** that became domain-irrelevant context, biasing both the *handling* of the scene (processed as an accident → less rigorously documented/searched) and later analyses as a favored suspect emerged. Honest status: **bias in the original work is now exposed** — context management would have helped — and the finding **excludes and includes no one**; it re-weights confidence in the *early* work, leaving us to trust most the findings (autopsy, instrumental chemistry) no expectation could steer.