30 June 2011

Fundamental Science Practices

How DO you do good science?  When I first joined the US Geological Survey (USGS) I was introduced into a powerful culture of high integrity science.  It came with an unusually stringent ethical code; I had to divest myself of some worthless mining stock that had been given to me.  As my branch chief told me, the only thing we have is our reputation as purveyors of extremely high-quality, unbiased science. Fail in that and we are nothing.

Since it came into existence with the Organic Act of 1879, the USGS has maintained comprehensive internal and external policy and procedures to ensure the quality, utility, and integrity of data, analyses, and scientific conclusions to maintain the reputation of USGS science for excellence and objectivity. In other words, certain specific high standards on how we must do science - from gathering data all the way to publishing it with an interpretation, and all the steps in between.  

In 2003, the USGS formalized these standards to ensure that the incoming biologists (yes, the USGS has a large biological discipline) as well as the geologists, hydrologists, and geophysicists all adhered to the same high standards.  A concept document was developed that outlined a set of science practices, philosophical premises, and operational principles that would serve as the foundation for USGS research and monitoring activities. Using the concept document and the best of our longstanding policy and procedures as the basis, in 2006, the USGS Fundamental Science Practices (FSP) policies were established.  Procedures related to applying these policies were developed, with the overall intent to ensure the quality and integrity of all future USGS science activities.

The FSP focuses on the operational principles related to planning and conducting data collection and research: how USGS science is carried out and how the resulting information products are reviewed, approved, edited, and released. The FSP are not designed to address the question of what work the USGS should do—this is addressed in many documents related to Bureau science planning.

A crucial part of the FSP is peer review of every science report or document: a minimum of two other disinterested scientists with expertise in the area of the report must do a critical review: did the authors think of this?  Did they follow a recognized valid procedure for that?  

Most important: will the final product - which must have both editing clean-up and Director's Sign-Off -  cause embarrassment to the USGS?  Director's Sign-Off is a read done by about 10 designated senior experts (not managers) who ensure that if there are political implications in a report - for instance a forecast of undiscovered hydrocarbon resources - that the USGS management gets a heads-up. Some have complained that this is in effect political interference, but in my experience this has never happened - at least in the USGS.

It seems like an admirable goal: nothing ever produced can be less than the very best, completely defensible - to maintain the reputation of the USGS. Historically this has been taken to such extremes that scientists have sat on a report, reworking the maps and the data, for years. The Best has sometimes become the enemy of the Good. Another potential problem is that risks are not often taken - there is a bias against going out into some completely new research direction.

For someone with an ego, finding all sorts of critical notes in red ink on your manuscript may be hard to take - until you see that what has been done by your peers is designed to protect you from embarrassment. 

The result is two-fold:

1. Every publication with the USGS logo on it is well-written, in impeccable English, and free of errors.  No fundamental assumptions have been missed, no logic bombs have slipped through.

2. If you are a USGS scientist, you can be justifiably proud of the scientific league you belong to.  


No comments:

Post a Comment