"Vital Statistics," by Claire Tilstone. Nature, 7 August 2003.
Statisticians question whether geneticists and molecular biologists have the statistical knowledge to analyze and interpret the vast amount of genomic data accessible now from microarrays (DNA chips). Nick Fisher, President of the Statistical Society of Australia, warns that "If the collection, analysis and interpretation of the data are flawed then it may not only be a waste of a valuable resource---we could draw faulty conclusions and potentially risk our health and environment." As the author puts it, "sloppy statistics could undermine the revolution promised by genomics and biotechnology." The article summarizes the challenges (good experimental design) and issues (interpreting microarray experiements, consequences of random "noise", variable data, etc.), and notes that "the high cost of the chips means that the number of repeated observations is usuallly very low." Fisher argues that research organizations should employ professionally accredited statisticians to help standardize experimental designs and to be involved before the data is collected, not after the experiment is done.
--- Annette Emerson