On Mon, 27 Jan 2003 13:03:12 +0200, Boris Liberman wrote: > Given the number of variables involved, it literally means zilch - > something like: There is a company called Pentax that happens to > produce some zooms, one of which apparently in my hands used in such > an such manner gives such and such results.
That's the core fallacy in most reviews of physical products. Rarely are statistical effects taken into account by the testing procedure. Sample-to-sample variation is not "weeded out" due to the extremely small number of samples tested together. TTYL, DougF KG4LMZ
