1. (a) Suppositions:

• There are aliens

• Aliens regularly visit our world

• Strange events occasionally happen that we should attribute to these aliens

• The aliens occasionally communicate telepathically with us, immobilize us and control our minds through some strange set of abilities they have

• Our thoughts can be screened

• We would not want our thoughts screened

• Aliens have taken some people prior to June 2000

• Having your thoughts screened, etc. has some bearing on whether or not one is prone to abduction (obviously, it is not clear exactly what this entails)

(b) One might want to leap upon the premise that aliens exist, but if that is construed broadly enough, it is probably wise to leave it alone. (We have no good reason to believe we are completely alone in the entire universe.) Whether they regularly visit our world or not is more questionable, since even unexplained UFO sightings do not confirm something of extraterrestiral origin. Whether we should attribute strange events to those creatures is also a matter worthy of contention, since their very existence is still open to debate. The same goes for their strange powers and their having made abductions prior to June 2000. One premise that should not come under serious attack from us though is that we would not want our thoughts screened; if such a process could take place, I imagine most of us would want to avoid it. As for the last, somewhat mysterious premise, I find it too unclear to accept, so I can only offer that as a reason to reject it.

(c) There is a clear post hoc fallacy in this piece. Let us assume that it is true that no one wearing the helmet since June of 2000 has been abducted. Give the author all the leeway he wants on this point. We could even assume that there were aliens abducting people all over the place, just for argument’s sake, and this would still be a post hoc fallacy. The problem is that the author has taken the lack of abductions among helmet-wearers since June as evidence of the causal effectiveness of the helmets. Of course, this is incredibly weak since the absence of abductions might be a mere coinidence; the aliens might not be taking new abductees at the moment, they may be avoiding people who wear the helmets for other reasons, or maybe, just maybe, there are no aliens abducting people at the moment. I hesitate to assign it any of the more specific sorts of post hoc errors that Walton mentions, since it is not even a positive correlation that is being cited. In this case, we can probably just say that it involves a post hoc error.

3. The problems worthy of mention would include the following:

(i) Data gathering: By using the Internet, the study selects out people with lower incomes and less formal education, who are less commonly regular Internet users, even though "blue collar" generally implies people in lower income brackets with less formal education. Thus, they are selecting out the very people they intend to study.

(ii) Size: 217 people is not a very large survey, especially given the other shortcomings of this survey. A much larger and more extensive poll would be required.

(iii) "People who visited our website": Note that this would suggest that the responses were received only from those who bothered to find the site and bothered to answer its questions. This would tend to select for people who strongly self-identified with some aspect of the page, which introduces the possibility of bias.

(iv) "Blue collar": The description does not include any way in which the different socioeconomic criteria were specified, only that there were three choices. Given that "wealthy" might carry some pejorative force to some people, those who answer the survey might be less disposed to offer this as an answer. In most cases, Americans in particular will describe themselves as "middle class" across very wide ranges of income. (Presumably this is in order to neither appear haughty and self-aggrandizing, nor to cast oneself as a member of an economic underclass.) Thus, self-identifying oneself with regard to these categories is not likely to be reliable.

(v) "Some problems sleeping at night": This is similar to the problem in (iii) in that it turns on a lack of clarity about what counts as a "problem" when it comes to sleep. I personally tend to sleep less than eight hours on a regular basis and feel fine, while others need more than eight hours on a regular basis. Again, what counts as a problem is not clear and so self-reporting is not expected to be reliable. (This problem also obviously occurs later in the description as well.)

(vi) The second survey: The simple problem here is that there is a mixing of the two surveys without any assurance that there is a similarity in the methods or that the methods are integrated into some larger plan. The description readily acknowledges that there is no socioeconomic self-identification, so it is not clear that any conclusion can be drawn based on any similarities there may be. (If anything, the fact that we get a similar number without that self-identification would suggest that the socioeconomic categories are irrelevant.)

(vii) The conclusion: Asserting that the sleepplessness is a result of financial stresses, whether or not these people really do feel such stresses, is too bold a conclusion to draw given the evidence. Other explanations are possible, but nothing about the data excludes them.

Ways to improve the study:

(1) If the study is to include socioeconomic categories, they should be laid out in terms of annual income and perhaps education. Assuming participants are fairly honest, this will at least allow for a more effective study of correlations between income and sleeplessness in terms that might actually pick out something of socioeconomic importance.

(2) Enlarge the study and actively seek participants rather than passively awaiting them.

(3) What counts as a "sleep problem" should be defined. There should be a way for participants to specify how much they normally sleep, how often they feel exhausted, whether there are irreuglarities to their sleep patterns (e.g. not at all during the week, lots on the weekend) and it should be made clear as part of the methodology (if not to the participants) what will count as a problem.

(4) If the aim of the study is to link sleeplessness with financial stresses, then the questions must ask whether there is any relation between the two and whether that relation is stronger for the different socioeconomic groups. Not everyone who is "blue collar" necessarily feels stressed about their finances and not every who feels stressed about their finances is necessarily "blue collar."

(5) Pursue the study through a means other than the Internet, since it still tends to select such a specific audience. (This is the one I feel the most hesitation in including, since this will not necessarily be true in the future. The ways we might think more appropriate - phone calls and mail service - were once the province of the affluent and well-educated as well.)

4. This episode of The Simpsons bears a great deal of similarity to other cases of procession to the mean that we have discussed in class. In this case, the important thing to bring out is that the selection of the players who appeared on the show was not an arbitrary process. Rather, players were selected on the basis of their name recognition with viewers, which was by and large a matter of how they had performed in recent years. It seems reasonable to assume that a star athlete’s performance would show the sort of distribution of success we described in class during any given stretch of time during a season, but one could also expect a similar pattern to emerge over the course of a whole career - some time developing as a player, several outstanding years, and then a waning before they retire. What got these atheletes on The Simpsons was generally a long and established career that made them household names, but this makes it much more likely that their careers would be on the wane. Throw in a few injuries and ignore the exceptions who continued or made a comeback (Ken Griffey, Jr. and Darryl Strawberry), and the correlation seems stronger than it really is. (It is notable that the episode was filmed and aired almost ten years ago, when Ken Griffey, Jr. was still in his early twenties and only beginning the prime of his career.)