Chapter 31 — Exercises

Work these without looking back at the chapter first; then check yourself. Items marked † have full worked solutions in the answers appendix. There are no answers in this file. Mix of recall, applied reasoning, evidence interpretation, "spot the overstatement," ethics, and a cold-case extension.

A. Recall and definitions

  1. Define cognitive bias in one sentence, and explain what the word systematic adds — i.e., why bias is not the same as random error.

  2. † Distinguish confirmation bias from contextual bias. Which one is usually the trigger and which is usually the engine, and what does that distinction mean in practice?

  3. Define domain-irrelevant information and give the counterfactual test the chapter uses to identify it. Then classify each of the following for a latent-print comparison as domain-relevant or domain-irrelevant: (a) the ridge detail of the latent; (b) the suspect's confession; (c) the substrate the print was lifted from; (d) the suspect's prior record; (e) an eyewitness identification.

  4. Define context management, blind analysis, and sequential unmasking, and state how the three relate (which is the umbrella, which is the strongest form, which is the refined practical procedure).

  5. Define the bias cascade in one sentence, and state the single most important consequence it has for how we should interpret several examiners "agreeing."

  6. What is the bias blind spot, and why does the chapter call it "lethal in forensics"?

  7. In sequential unmasking, what must the examiner do before the reference sample is revealed, and why is the written, time-stamped form of that step essential?

  8. Name the two organizational "pillars" the chapter says make context management real rather than aspirational (one is a role, one is a documentation requirement).

B. Applied reasoning

  1. † An examiner is handed a poor partial latent, a suspect's exemplar, and a note: "This is the guy — he confessed; just confirm the print." Identify (a) the domain-irrelevant information present, (b) the specific word in the note the chapter calls "an alarm," and (c) two distinct ways this setup biases the comparison.

  2. Explain mechanically how an unexplained discrepancy (a ridge ending in the exemplar but not the latent) is handled differently by an examiner who expects a match versus one who has no expectation. Which bias is doing the work, and why does the same physical feature get opposite treatment?

  3. A laboratory proudly notes that every identification is "verified by a second examiner." Under what condition is this verification worthless as a check on bias, and under what condition does it become genuine corroboration?

  4. † A detective says, "If the analyst is a true professional, they can be told the whole case and still be objective." Using §31.1, explain precisely why this confuses the issue, and why an examiner's sincerity is not a safeguard.

  5. Walk through the three stages of sequential unmasking for a firearms/toolmark comparison (a questioned bullet vs. a test-fired bullet from a suspect's gun). State what is revealed at each stage and what is withheld, and identify the most-biasing information that should never reach the bench.

  6. Explain why instrumental methods (e.g., a GC-MS spectrum, Chapter 23) need context management least, and why a DNA mixture interpretation (Chapters 8–9) needs it more than a single-source DNA profile. What property is doing the explaining?

  7. The chapter says bias is a structural problem with a structural solution. Explain what that means, and why it implies that "hire better people / train them to be objective" is the wrong fix.

C. Evidence interpretation

  1. † Re-read Figure 31.2 ("Reading the same latent two ways"). State precisely what the figure claims the Mayfield-style error was and what it claims the error was not. Why is the distinction between those two readings the whole point of the chapter?

  2. Re-read Figure 31.1 (the bias cascade). Trace a single piece of domain-irrelevant information ("the suspect confessed") from its entry point to three different places it ends up contaminating. Then state where the cascade "feeds back" into the investigation and why that feedback loop is so hard to unwind.

  3. An expert testifies with total confidence to a fingerprint identification. From the witness stand, a contaminated conclusion and a clean one "sound exactly as confident." Using the chapter, explain why this is true and what question a juror or attorney must therefore ask instead of "is the expert sure?"

  4. A reviewer signs off on a colleague's comparison after being told, "Examiner 1 already called this an identification — please confirm." Name the bias(es) at work and rewrite the instruction so the review becomes a genuine blind verification.

D. Spot the overstatement / the comforting ritual

  1. † A laboratory director states: "We have fully addressed cognitive bias — we now require peer review and technical verification of all conclusions." Identify the specific way this claim can be false even though peer review and verification are happening, and state the one-word property that distinguishes a real fix from a comforting ritual.

  2. An attorney argues, "The Dror experiments prove that fingerprint examiners get the answer wrong most of the time." Identify the overstatement: what did the experiments actually demonstrate, and what did they not demonstrate? Rewrite the claim honestly.

  3. A prosecutor says, "Four separate experts independently agreed it was a match — that's overwhelming corroboration." Under what hidden condition is this sentence misleading, and what is the honest version?

  4. A trainer tells new examiners: "Bias is only a risk for sloppy or dishonest analysts; if you do careful work, you're fine." Identify every error in this sentence using §31.1–31.2.

E. Ethics and reasoning

  1. † A small two-person laboratory inside a police department cannot easily implement blind verification (there is rarely a second free examiner) and feels pressure from detectives to "help make the case." Discuss the ethical tension honestly: name three real obstacles from §31.6, and propose at least one partial safeguard the lab could adopt even under these constraints.

  2. You are an examiner and you realize, mid-comparison, that you were told the suspect confessed before you began. What is the honest thing to do, and how should it be documented? Why is concealing that you were exposed to the context worse than disclosing it?

  3. Is it ethical to testify to a confident "identification" reached through a workflow you know was not blinded, without disclosing that fact? Argue the position the chapter's voice would take, connecting it to the duty to the court (Chapter 30).

  4. The chapter insists the threat is "not bad analysts, and not dishonest ones." Why is this framing ethically important — both for getting examiners to accept reform and for fairly judging past errors like Mayfield?

F. Synthesis and validity spectrum

  1. † Place these on a ranking of how much each needs context management (most → least), and justify each: a bite-mark comparison (Chapter 16); a single-source DNA profile (Chapter 7); a latent-print comparison (Chapter 14); a GC-MS drug identification (Chapter 23); a bloodstain-pattern interpretation (Chapter 10). Then explain how this "bias-vulnerability" ranking relates to, but is not identical to, the NAS/PCAST validity-spectrum ranking.

  2. Explain how this chapter reinterprets a method taught earlier. Pick one (fingerprints, toolmarks, bloodstain analysis, or DNA mixtures) and state specifically what new vulnerability the chapter exposes that its own chapter could not fully convey at the time.

  3. In one paragraph, explain how this chapter advances at least two of the book's four themes (exclusion over proof; the validity spectrum; cognitive bias; the CSI effect cutting both ways). Name which themes and how. (This is the home chapter for one of them — say which, and why "home chapter" is the right description.)

G. Cold-case extension

  1. † Cold Case. The chapter's Case File concludes that "bias in the original work is now exposed." Write the entry you would add to the Mill Creek workbook (Appendix I). State (a) what the contaminating expectation was, (b) how it could have propagated (the cascade), (c) why this finding excludes and includes no one, and (d) which findings in the file you would now trust most because no expectation could have steered them — and why.

  2. Cold Case extension. Re-read the cold case's gas-can latent outcome (Chapter 14, inconclusive). Suppose you are designing the lab policy that produced that good outcome. Write a three-stage sequential-unmasking protocol for how that latent should have been handled, naming what context the bench examiner is given and what the case manager withholds.

  3. Cold Case, integrative. The "accidental fire" frame biased the handling of the scene (Chapter 2), not just the lab analyses. Explain how an early expectation about the conclusion (accident) can contaminate the collection and documentation of evidence — and why that kind of contamination is, in some ways, harder to fix later than a biased comparison. Name one finding in the file that is robust to this contamination and say why.

H. Short writing

  1. In 150–200 words, explain to a juror why a forensic expert's confidence is not, by itself, evidence that the conclusion is correct — and what the juror should ask instead. Use plain language and at least one idea from §31.3.

  2. † In 150–200 words, make the strongest honest case for sequential unmasking to a skeptical veteran examiner who says, "I've testified four hundred times and never needed to be kept blind to the case facts." Address the bias blind spot, the difference between sincerity and protection, and what the examiner gains (not just gives up) by working blind.