Key to Chapter 16 Review Questions

§16.1 Introduction

  1. (a) Without accurate knowledge of cause, 'the effort which have been made to oppose [the problem will have] a contrary effect' (paraphrasing Snow's quote, p. 289).  
    (b) When discovery of effective preventive measures pre-date identification of the definitive cause.
  2. Causal inference is the process of deriving cause-and-effect conclusions from fact and knowledge.
  3. No, but a statement of proof can be strong and overwhelming.
  4. True
  5. The two types of epidemiologic decisions are (1) those having to do with scientific hypotheses and (2) those having to do with public health actions. The former requires rigorous skepticism. The later may  require making a reasonable choice based on available information.
  6. See Table 16.1, p. 290.
  7. Induction is the process of deriving general law from particular observations.  The Problem of Induction is the  philosophical quandary that observed sequences do not prove cause and effect (post hoc propter hoc). 
  8. True. Refutationists believe that a theory is not scientific unless it is falsifiable. (Note that not all scientists are Refutationists.)
  9. Because the next swan you see may not be white. See "The Problem of Induction." 
  10. No number of observations can prove a hypothesis, but one strong disproof can nail it shut. 

§16.2 Advisory Committee to the Surgeon General  on Smoking and Health

  1. 1964 
  2. False, of course
  3. True 
  4. Consistency; Strength; Specificity; Temporality; Biological coherence

§16.3 Hill's Framework

  1. Ratio measures, like the RR, SMR, and OR
  2. Strong associations are less likely to be explained by confounding. (See the comment by Hill in the middle of page 294 for an elegant discussion of strong vs. weak associations.)
  3. No. There are many occasions in health when slight associations are true.
  4. No. They may all be "incorrect".  
  5. True. The cause must always precede the effect. Proper temporality is (perhaps) the only required criteria.
  6. See Fig 16.2: merely replace T. canis with "environmental lead".
  7. Alcohol consumption and cardiovascular disease risk (i.e., moderate levels decrease CVD risk; high levels increase CVD risk).
  8. Plausibility
  9. Coherence holds that all sources of evidence "stick together". Plausibility holds that relations can be explained by current knowledge.
  10. In vitro, in vivo animal models, in vivo physiologic studies, community/clinical/field trials.
  11. Analogy
  12. Biological gradient
  13. Consistency, Strength, Specificity, Temporality, Biological Gradient, Plausibility, Coherence, Experimentation, Analogy
  14. (1) = e, (2) = f, (3) = a, (4) = b, (5) = d