On Thursday 20 October 2005 12:03, Juan Antonio Maciá Agulló wrote:
Thank you very much Bruce for your explanation. As you have commented, I have high correlations between S02 and sigma^2, I can not eliminate them by changing kw, so... what can I do?
The answer might very well be that you have to live with it. These correlations will be reflected in the size of the uncertainty. In many cases it is possible to change some kind of extrinsic parameter and invoke some knowledge of how your parameters change when thatr extrinsic parameter is changed. An example of an extrinsic parameter is temperature. We presume that S02 is constant in temperature and that sigma^2 behaves like a single-frequency oscillator with temperature. If you are able to make a sequence of measurements as a function of temperature, the variations in the data combined with your understanding of how the data should behave can be very useful in determining parameters with more confidence. Of course, there are many situations where it is not possible to take measurements as a function of temperature (or presure or whatever). In that case we have to do the best we can and acknowledge that there is some systematic uncertainty in the system. In that case, one must be honest about the uncertainties when reporting results. B -- Bruce Ravel ---------------------------------------------- bravel@anl.gov Molecular Environmental Science Group, Building 203, Room E-165 MRCAT, Sector 10, Advance Photon Source, Building 433, Room B007 Argonne National Laboratory phone and voice mail: (1) 630 252 5033 Argonne IL 60439, USA fax: (1) 630 252 9793 My homepage: http://cars9.uchicago.edu/~ravel EXAFS software: http://cars9.uchicago.edu/~ravel/software/exafs/