Course discussion blog for "How to Lie with Statistics: Uses and Misuses of Numbers in Argument", a 300-level Honors course at the University of New Mexico. Anyone can read this blog, but only class members can post.
Monday, February 4, 2013
How We Lie
In his second chapter "American Polygeny and Craniometry before Darwin" Gould states that "expectation is a powerful guide to action," and this quote has stuck with me throughout our reading of the book (97). As the title of this course suggests, one questions that we are continually faced with is how statistics are or are not used to lie. As Eagleman says in his video "The Brain and the Law," we still have a very murky understanding of what a lie actually is, for someone may be "telling the truth but factually wrong." I think this is an especially important point to consider while reading Gould in respect to the many statisticians and scientists he presents. For example, Gould states that Broca "did not fudge numbers; he merely selected among them or interpreted his way around them to favorable conclusions" (119). Such a statement begs the question of whether or not Broca "lied" or, like Eagleman suggests, is telling some sort of truth while still factually wrong. Thus I think it is important to consider the idea that "theories are built upon the interpretation of numbers, and interpreters are often trapped by their own rhetoric" (106). While "numbers don't lie" is an oft heard colloquialism it seems apropos to keep Gould's cautionary words about numbers in mind when considering our recent readings of chapters 5 &6, for those detail individuals like Yerkes, Binet, and Cyril Burt who attempted to assign a precise number to an intangible quantity, intelligence. While their numbers may not be factually "wrong" within the system's them themselves have created, they are still a product of their interpreters.
Labels:
TopicBlogPost
Subscribe to:
Post Comments (Atom)
Julia,
ReplyDeleteYou bring up some really good points and I too was struck by the accuracy of Gould's quote, "expectation is a powerful guide to action". When thinking about expectations guiding action however, I think it is important to distinguish between conscious and unconscious expectations. While there are certainly instances in which a researcher is aware of their bias and uses said bias to interpret their data, I believe that there are also cases where researchers are unaware of their biases. This poses a major threat to the validity of scientific findings. How can a researcher combat their own biases if they are unaware that they exist in the first place?