Tuesday, February 12, 2008

Micro Statistics Tutorial 07: Lessons from honey bees

Here is some correspondence in Science 30 years ago in response to a graph and statistics in a paper by Roubik (Science 1978: 201;1030).

Which prompted Robert Hazen to respond (Science 1978 202:823):

I liked Hazen's response and the subsequent response by Roubik (not shown here). At least this is proper science.

When things are visible and transparent they can be discussed. Despite the obvious over-fitting of data, the Roubik's paper was open and transparent (unlike a lot of the pharmaceutical literature). At least the data were shown.

Misleading graphs formed part of the attempt by Procter and Gamble Pharmaceuticals (P&G) to interfere with scientific evidence about their osteoporosis drug Actonel. The problem was that it was not possible for anyone to know this (including the authors and scientists involved) because there was no data. This was compounded by further spinning of words in the embarrassing admission of guilt over the first of the three manuscripts. Watch this space.

Scientists may disagree about the presentation of data, but there can be no legitimate debate when that data is not available, even to the authors.

See here for Collated Micro-Statistics Tutorials

Earlier|Later|Main Page

No comments: