1. (4 pts) Key code, ID #, and name. Fill in (A B) to indicate your key for this version of the exam. Be sure your name and ID number are correctly bubbled in on the scantron and that you have signed this exam copy. See the page displayed on the big screen for details of where your ID number goes.
Intrinsic Difficulties
2. (5 pts) Which of the following are true? MTF
A) The coin flip demo during the intrinsic difficulties lecture was used to illustrate how rare events could be generated merely through sampling error. This demo was not done in 2008; it was used to illustrate a method of biasing a study, nothing to do with rare events.
B) The flash powder demonstration illustrated that a fundamentally different outcome could come from the combination of two things than from either alone; this was a type of interaction or complexity. Yes, the flash powder is a mix of two chemicals, neither of which alone would flash.
C) ‘Intrinsic difficulties’ refers to types of problems for which the scientific method cannot ultimately obtain answers, hence we need to find some other way of addressing those problems. No, intrinsic difficulties CAN be solved with the SM, they are just more challenging than most.
D) The theme given that unites intrinsic difficulties is that they all stem from problems in evaluation (as opposed to data, models, or revision). No, most difficulties lie at the level of the SM itself, not in any one component. For example, ‘time lags’ slow the entire cycle of the SM.
E) We noted that intrinsic difficulties are expected to become less important in the future, as societies develop a better understanding of the scientific method. No, the fact that the difficulties are ‘intrinsic’ means that the SM will always have difficulties. “Humans’ might become less of a problem if future generations decide to relax safeguards to protect us, but the others are out of our control
3-7. (3pts each) In the following problem descriptions, which intrinsic difficulties are indicated? Do not infer more than is described. The underlined part of the question illustrates the problem that needs to be addressed by your answer. Use the following options for your answers: (one per question)
|
(A) Time lags |
(B) Rare events |
(C) Humans make difficult subjects |
(D) Interactions/ complexity |
(E) None |
3. The drug DES was used in the 1940s into the 1970s to suppress miscarriage. It was discovered that the drug caused elevated cancer rates. Which intrinsic difficulty explains why the drug was used for 30+ years before it was withdrawn? (A) (B) (C) (D) (E) one only Time lags. This one required that you rely on notes from lecture (and maybe the book). The problem was that the cancer happened in the offspring of the mothers who took the drug, and it arose in the daughters in their 20s and 30s.
4. The FDA has recently approved several drugs that were subsequently withdrawn because they created problems for some patients (e.g., Baycol and Vioxx). The clinical trials had missed those problems because they arose in 1 of every 10,000 patients. Clinical trials only use up to 3,000 patients. Which difficulty explains why the drug passed clinical trials yet was eventually found to cause problems? (A) (B) (C) (D) (E) one only Rare events. If something occurs in only 1 of 10,000, a trial with 3.000 will usually miss it completely.
5. The FAA, which oversees airline safety, has recently instituted a model of accidents (crashes) that uses data on near misses. Near misses are perhaps a poor model of crashes, but they are much more frequent than actual crashes. Airline safety has gotten to the point that there are not enough crashes to obtain adequate data on their causes. What difficulty explains why the FAA has resorted to this alternative model of crashes? (A) (B) (C) (D) (E) one only Rare events, from the ‘are not enough crashes to obtain adequate data on their causes’.
6. Superconductors are materials (metals) that have almost no resistance to electricity. Superconducting materials were long sought but only created maybe two decades ago. They consist of combinations of odd metals, and none of the metals alone have superconducting materials so there was no way to predict how to make them. What difficulty underlies the long time it took to discover how to make superconductors? (A) (B) (C) (D) (E) one only (A = time lags; B = rare; C = humans; D = interactions; E = none) Interactions, from the ‘none of the metals alone have superconducting’ ....
7. Which difficulty is illustrated by the unwillingness of parents to accept the results from FC experiments (in the second part of the FC video)? (A) (B) (C) (D) (E) one only None. The unwillingness was due to conflict, not intrinsic difficulties.
8. (4 pts) Clinical trials are required for FDA approval of new drugs and some food products. Which points about clinical trials are true? MTF Covered on 3rd exam in 2008. Pretty much just factual questions.
A) There are 3 different levels (phases) that must be undertaken before approval (I, II, III). Yes. Phase IV applies only when an approved drug is tested for a new application.
B) The different phases all involve approximately the same numbers of patients. No, the numbers increase with phase, so phase III is the largest.
C) In terms of safety to the participants, Phase I is safest. This is a test of toxicity and is the least safe.
D) The typical phase III involves a randomized, double-blind procedure. If you participate in a phase III trial, you have only a 50% chance of being assigned the drug being tested. Yes, as read in class.
Biological Determinism (BD, for short)
9. (4pts) Which of the following are true, as covered in class? MTF
A) Alternative ways of describing the question of biological determinism is “nature vs. nurture” or “is it in our genes?” Yes, as given in class.
B) As we discussed, a modern controversy over the question of BD concerns genetically modified organisms and whether their genomes are likely to be harmful to society. May sound good, but we never discussed GM organisms. Furthermore, BD is about humans.
C) BD is relevant to IQ and race, gender roles in society, eugenics, and crime. Yes. We listed all 4.
D) The castration of convicted rapists in exchange for their freedom, which was mentioned in class, assumes that their raping behavior is biologically determined, and thus is largely genetic. Mostly correct, except the last part. If the behavior was genetic, castration would not change the behavior because it does not change the genotype.
10. (4pts) Which of the following are true, as covered in class? MTF
A) State laws in the U.S. during the 1900s allowed the castration of individuals judged to be feebleminded. Yes. We even showed a picture of someone whose forcible castration was upheld by the Supreme Ct.
B) The castration of individuals deemed to be ‘inferior’ as a way of preventing them from having children, was motivated by the presumption that their ‘inferiority’ had a genetic basis. Yes, this is the essence of ‘eugenics.’
C) The German Nazi regime of the 1930s and 1940s incorporated an extreme form of biological determinism, castrating or killing many of its citizens in attempting to create a master race. Yes, as gone over in the pamphlet shown in class.
D) Although parents affect their children’s behaviors through non-genetic means, such as by education and by setting examples, modern genetics methods have resolved this problem so that inheritance of human behaviors can be assessed confidently from parent-offspring resemblance. No, you still have the problem that the home environment may influence the person’s behavior, IQ, etc.
11. (5pts) Sexual preference in humans: which are true? MTF
A) Studies of inheritance, based on comparisons of identical twins to full sibs, suggest the lack of a genetic/inherited basis to sexual preference. No, we gave numbers for identical twins, full sib brothers, and unrelated brothers that suggest a partial genetic basis to SP.
B) The relevance of biological determinism to sexual preference is whether sexual preference is a choice (and possibly learned) or is something that a person is born with. In the past, at least, a popular public belief has been that homosexual preference was learned. Class mentioned a 1970s survey in which over 40% of those surveyed thought gay behavior was learned.
C) Themes for this topic included ‘humans make difficult subjects’ and ‘correlations.’ Both were mentioned as themes.
D) An anatomical difference between heterosexual and gay men has been observed in some studies on the finger length ratio (index to 4th finger). Gay men have finger length ratios more similar to those of women than those of heterosexual men. First sentence is true. Second is not – this is one trait in which gay men appear to be overmasculinized.
E) A correlation has been observed between sexual preference of a man and the number of older brothers he has, each older brother increasing the chance that he is gay. Yes. 3-4 older brothers doubles the chance he is gay.
F) Sexual preference in humans is also known as with sexual identity. No, these are two separate dimensions of human sexuality.
12 (5pts) LeVay’s study on the brain of gay and heterosexual men: which are true? MTF
A) This study demonstrated a correlation but not causation. Yes. Indeed it is almost impossible to demonstrate causation in humans. He did observe a correlation.
B) LeVay looked for a difference in the size of a brain region between gay men and heterosexual men; this region contained several ‘nuclei,’ each nucleus was about the size of a cherry No. Each nucleus was mentioned as being microscopic.
C) LeVay observed a size difference in one brain nucleus that was associated with sexual preference in men. Heterosexual men had a larger nucleus than gay men, whereas the size in gay men was similar to that of women. Yes.
D) Many of the brains used in the study were from AIDS deaths. It was not possible to rule out the nucleus size difference between gay and heterosexual men as being due to AIDS. Wrong. It was indeed possible to control for AIDS deaths between het. and gay men.
E) In the brain shown in class, the region studied by LeVay (known as the anterior hypothalamus) was on the outside front of the brain. No, it was near the middle on the inside.
Recall that ToC requires (i) a group resource, and (ii) selfish interests work against group benefit.
13 (4pts) Each of the following options contrasts two properties of a vaccine, disease, or population. Mark those in which the first property (italicized) is more likely to result in a ToC conflict than the second property. MTF
A) herd immunity exists for the disease/ herd immunity is absent herd immunity is a type of group benefit, a requirement for ToC, so the first is more likely to result in ToC.
B) individuals get vaccinated for selfish reasons/ individuals avoid vaccination for selfish reasons ToC needs the selfish interests to go against group interests. Thus, the second is more likely to result in ToC.
C) the infectious agent spreads from person to person/ the agent does not spread person-person This is pretty much the same as (A). You can have herd immunity only if the agent spreads from person to person. Thus the first property is more apt to create ToC.
D) the vaccine prevents the individual from getting infected/ the vaccine does not prevent an individual from getting infected but does prevent transmission by the infected individual Somewhat similar to (B). If the vaccine does not prevent the person from getting infected, then the only reason to get vaccinated is to protect others. So the selfish interests are to avoid vaccination in the second case, which underlies ToC.
14 (6pts) Which of the following describes Toc conflict or outcome? Recall that ToC is not limited to humans. MTF
A) A human body is a population of about 1013 cells. Through regulated growth, all cell types are maintained at requisite levels and contribute to the survival of their own lineages as well as to the survival of all other cell types (the body). A cancer is a population of cells that grows abnormally fast. When a cancer lineage first arises, it is rare. But its numbers increase disproportionately because of their high growth rates. By growing fast, cancer cells ‘selfishly’ increase in the body. If untreated, their numbers eventually become so high that the body ceases to function, and all cell types die. Yes, this one is a non-human type of ToC – the selfish interests of the cancer cell killing the body.
B) Six farmers form a cooperative to manage their separate lands. During a long drought, it becomes necessary to irrigate. Collectively, they make a decision to pump water from the aquifer faster than it is replenished, and eventually, it runs dry. All six farmers suffer financially in future years because of the dry aquifer. No. There is a group resource, but no selfish cause of the demise of the resource.
C) 20 students form a common bank account, because they anticipate higher interest rates with a collective account than with individual accounts, presuming that they will each maintain reasonable balances. Each student puts in the same amount of money up front. There are limits on each account so that each student can withdraw no more money than each put in, so the account is protected from students taking out the money of others. The interest is shared equally, however, regardless of the individual account values. In the long term, everyone keeps most of their money in separate accounts because each gets all the interest due them, and the balance is low in the joint account that no interest is generated. Yes, the group resource is the interest. Selfish interests cause its demise.
D) A rancher owns 6 different properties, all adjacent (maintained as separate properties for tax purposes). There is a common aquifer spanning the 6 properties. 4 properties are used for grazing cattle, 2 for crops. The rancher’s decision to irrigate the crops eventually causes the aquifer to dry up, at which point it is not only no longer feasible to grow crops on the two properties, but it is also no longer possible to keep cattle on the other 4. No. There is no group ownership.
15. (6 pts) Which of the following are true about conflict in general, not just “tragedy of the common.?” Some options require distinguishing conflict from bias. MTF
A) All types of conflict we considered are different versions of “tragedy of the common” conflicts. No, ToC was specifically group vs. individual. Another was individual vs. individual.
B) Conflict between the defense and prosecution over evidence in a criminal trial leads to both sides treating evidence, especially evidence that involves uncertainty, differently than is scientifically proper. Yes, as in lecture and book.
C) The effect of Lysenko on Soviet genetics was given as an example of conflict due to political factors dominating scientific ones. Yes.
D) It was suggested that the social acceptance of science in the U.S. has advanced to the point that political factors no longer subvert science. Certainly not.
E) Bias usually appears in the evaluation stage of a study. Conflict usually appears in the design. No. conflict is at the level of goals. Bias can occur at any level.
F) Conflict refers to different people/institutions having different goals. Bias refers to a way of distorting a study or its conclusions away from “truth.” Yes, and would help you answer (E) above.
16 (4pts) Which are true of the second FC video and the related discussion in class? MTF
A) Several instances were shown in which individuals refused to doubt the validity of FC. For those FC-accepting individuals interviewed (some parents, and the administrator Doug Bicklen) it was apparent from the video that factors other than seeking the truth could be affecting their views. Yes, the parents did not want to give up the hope that their child could communicate; Biklen was portrayed as founding an Institute.
B) Bicklen raised objections to the tests of FC on the grounds that the testing environment was intimidating. The controls used in the tests of FC supported his concern. No, the controls could have been used to argue against him.
C) One set of parents was shown who accepted the evidence that FC was bogus. They had a strong incentive to doubt its validity, because FC had been used against them. Yes, FC had been used as the basis of an allegation of sexual abuse.
D) According to the video, FC had been widely accepted before it had been tested. Yes.
E) Syracuse University still maintains an institute for Facilitated Communication. It does, as of 2008, but we did not show this in class in 2008, only in 2007.
17 (4pts) Courts and conflict/bias (DNA) MTF
A) There is a built-in conflict between the defense and prosecution. Labs providing the analyses are largely free of conflict with either defense or prosecution. No, we suggested that the labs used by the prosecution are in conflict with the defense.
B) There are many legal procedures that have been used to thwart legitimate scientific analysis of evidence in the courts. These include the use of known sympathetic experts and the prosecution’s failure to include testing procedures that might benefit the suspect. Yes, we listed these and more.
C) The prosecution is governed by strict rules of conduct with respect to its interactions with the various parties in a trial. Thus, for example, the prosecution would never harass defense witnesses outside of court. No, we mentioned that prosecution attorneys have indeed harassed defense witnesses outside of court.
D) The legal system deals with uncertainty in a fundamentally different way than does ‘science.’ In science, uncertainty is an acceptable conclusion. In a trial, a decision about guilt must usually be made, which is not compatible with uncertainty. Thus uncertainty is inflated by the defense to argue in favor of acquittal but is deemphasized by the prosecution. Yes. This option is pretty much the same as 15(B) above.
18 (6pts) Bias can often be recognized in the form of non-scientific arguments. Which in the following list were given as examples of non-scientific arguments indicative of bias? Do not include options that describe how bias is created. MTF Must be an argument and reflect bias. Only the wrong answers are explained in red.
A) appeal to authority
B) control the null model Not an argument
C) throw out unwanted results Not an argument
D) character assassination of opposition
E) identify trivial flaws in an opponent’s model
F) identify major flaws in an opponent’s model highlighting a major flaw would not indicate bias
G) defend an unfalsifiable model
H) defend a falsifiable model No, the model SHOULD be falsifiable.
I) assay for a narrow spectrum of unlikely results Not necessarily an argument; this option should have been graded either way.
J) assay for a broad spectrum of results Not indicative of bias.
19-23 (3pts each) Some other non-scientific arguments (other than those in the preceding question) are given in A-E below. Which of those apply to the following questions? One answer only.
|
(A) use anecdotes and post hoc observations |
(B) Either-or arguments |
(C) Build causation from correlation |
(D) Require refutation of all alternatives |
(E) Refusal to admit error |
19. Doug Bicklen’s claim that it does not matter how many failed attempts there are when trying to show whether FC works. (A) (B) (C) (D) (E) (one only) A clear example of refusal to admit error
20. A student claims that a test debunking horoscope accuracy has not convinced him that there is “nothing to” horoscopes, and thus that he still believes in them. (A) (B) (C) (D) (E) (one only) This is tantamount to requiring that all possible models of horoscopes be rejected, D.
21. A parent arguing that FC works for them, because they can recall one instance when the typing seemed to make sense. (A) (B) (C) (D) (E) (one only) This is a clear example of an anecdote.
22. Creationists argue that the inability of the theory of evolution to explain some particular observation means that the theory of special creation must therefore be correct. (A) (B) (C) (D) (E) (one only) Either-or, means that if A isn’t correct, then B must be correct.
23. A lawyer argues that his/her client’s suicide attempt was caused by the company’s antidepressant, because there is a clear association of suicides with antidepressants. (A) (B) (C) (D) (E) (one only) This option describes a correlation in the second part of the sentence, and infers causation in the first half. So it is an example of building causation from correlation.
24-32. (2 pts each) Bias can be introduced at several levels in a study, as given below. Which of those apply to the following questions? One only for each question
|
(A) Before the design and conduct |
(B) In the design and conduct |
(C) In evaluation and presentation |
(D) None |
24. Non-random assignments (A) (B) (C) (D) This would happen in the design and conduct.
25. Small samples to avoid unwanted, rare outcomes (A) (B) (C) (D) This would in the design
26. Control the null model (A) (B) (C) (D) This operates BEFORE the design stage.
27. Statistical analysis (A) (B) (C) (D) The analysis itself is done after the design and conduct.
28. Assay for a narrow spectrum of unlikely results (A) (B) (C) (D) This would be in the design and conduct – you decide what you will assay for before you gather the data.
The following cases refer to abuses of the drug industry, as in the book:
29. Tell only part of the story (A) (B) (C) (D) This will happen when the study is done, C.
30. Comparisons confounded by dose differences (A) (B) (C) (D) Could be clearer, but dose differences would be part of the design.
31. Testing the wrong age group (A) (B) (C) (D) The age group studied would be part of the design.
32. Using ghostwriters to put a good spin on the study (A) (B) (C) (D) ghostwriters are employed when the study is done, C