Science for Business, Law and Journalism.Chapter 10. Reducing the error: A template for ideal data.
Grandjean, P. et al. 1998. Cognitive performance of children prenatally exposed to "safe" levels of methylmercury. Environmental Research, Section A 77: 165-177. Note: You have already seen this paper. Once again, do not read it from front to back. Rather, read it with a view towards answering the questions posed below. Whereas we discussed models, impediments and errors previously, in this class our focus will be on different good research design.
Schober, S.E. et al. 2003. Blood mercury levels in US children and women of childbearing age, 1999-2000. JAMA 289: 1667-1674.
Atomic Absorption Spectrophotometry A short introduction from the Royal Academy of Chemistry.
Background material on spectrophotometry
1. We will march through the five features of good experimental design (explicit protocol, replication, randomization, standards and blind), asking whether each is present in the Grandjean et al. and Schober et al. studies. For those absent, we will ask what problems this creates.
2. Work through each of the 5 features of ideal research design, explaining how each can be used to reduce error, and the type(s) of error each of these 5 features of good experimental design reduces. For example:
What types of error are we attempting to reduce by using an explicit protocol? Answer: Primarily bias - explain this answer. What types of error are we attempting to reduce through the use of replication?
What features of the ideal data template are most helpful in reducing human error?
And so forth. Hint for the "and so forth": Take a sheet of paper, and lay out the 4 types of error and 5 features of ideal data in two lists, side by side. With this framework, it is easy to generate many questions like the examples posed immediately above.
The following questions duplicate those posed immediately above. However, they are sufficiently important that it is worth highlighting them.
3. Explain how a large number of replicates will help prevent sampling error.
4. Explain how a blind experimental design will prevent unintentional bias.
5. How do lawyers use the term bias? How does the legal use of bias differ from its use in science?
