This file is http://www.nevis.columbia.edu/~seligman/ccfr/SYSTEMATICS

This is a description of the systematic-error files that are
in the seligman/ sub-directories.  For a detailed description of
each systematic error, see William Seligman's thesis.  For an
overview of the contents of the sub-directories, see
/~seligman/ccfr/READ_ME.


In the systematic error analysis, the entire structure function
(SF) extraction is repeated for each systematic quantity to be
varied.  The files used in each systematic extraction are 
distinguished by their suffixes.  The meaning of the different 
suffixes are as follows:

-base (or no suffix)
        The file was extracted for the main (or base) SF.  No
        systematic quantity was varied.  This file represents
        the final result from the analysis.

-e744-hadscl+
-e744-hadscl-
-e770-hadscl+
-e770-hadscl-
        The hadron energy scale was increased (decreased) by 1%,
        i.e., E(had) = E(had) * 1.01 (.99)
        This systematic is computed separately for experiments
        E744 and E770.

-ehad+
-ehad-
        The hadron energy was increased (decreased) by 100 MeV,
        i.e., E(had) = E(had) +(-) 100 MeV.

-muscal+
-muscal-
        The muon energy scale was increased (decreased) by 1%,
        i.e., E(mu) = E(mu) * 1.01 (.99)

-emu+
-emu-
        The muon energy was increased (decreased) by 100 MeV,
        i.e., E(mu) = E(mu) +(-) 100 MeV.

-e744-ba-nu+
-e744-ba-nu-
-e744-ba-anu+
-e744-ba-anu-
-e770-ba-nu+
-e770-ba-nu-
-e770-ba-anu+
-e770-ba-anu-
        The value of B/A derived from the flux extraction was
        increased (decreased) by one sigma as determined by a fit
        to the low-Ehad data.  This systematic is computed separately
        for neutrinos and anti-neutrinos, and for experiments
        E744 and E770.


-flxadj+
        The flux was _not_ adjusted so that sigma/E would be linear.
        If there is a "flxadj-" file, it was not extracted directly,
        but computed by looking at the difference between
        "-base" and "-flxadj+".

-sigmax+
-sigmax-
        The overall normalization for the neutrino-iron cross-section
        was increased (decreased) by 2.1%.

-sigrat+
-sigrat-
        The ratio of the anti-neutrino to neutrino cross-sections
        sigma(nubar)/sigma(nu) was varied by 1.4%.

-r+
-r-
        The value of R used in the analysis was increased (decreased)
        by 15%.

-model+
        An alternate Buras-Gaemers physics model was used to extract the
        contents of this file. See the Seligman thesis for more
        information.

The following systematic parameters are all derived from the leading-
order dimuon analysis of Steve Rabinowitz (SAR).  Since they have a small
effect on the structure functions, only the effect of increasing the
parameter was extracted.  If a "-" file exists (e.g., "-alpha-"), it was
derived by comparing the "+" file to the "-base" file.

-alpha+
        The value of alpha (the strange sea shape parameter) was
        increased by one sigma as defined in the SAR analysis.

-brc+
        The value of charm branching ratio was increased by one
        sigma as defined in the SAR analysis.

-cmass+
        The value of the mass of the charm quark was increased by
        one sigma as defined in the SAR analysis.

-kappa+
        The value of kappa (the strange sea normalization parameter)
        was increased by one sigma as defined in the SAR analysis.



If a given systematic quantity has no file associated with it, it
means that the quantity is not relevant the contents of that file.
For example, there is a file named FLXM744-hadscl+, since varying the
hadron energy scale has an effect on the flux, but there is no
FLXM744-sigmax+ file, since the flux is not normalized to the
cross-section in our flux extraction (that normalization takes place in the
structure-function extraction).


MODEL CORRECTIONS
-----------------
-no-dxf3
-no-iso
-no-rad
-no-cmass
-no-propag
-no-corr
        These files contain the results of extracting the structure
        functions without one or more of the physics model corrections.  

        The corrections included in these files are the "delta xF3
        correction" (xF3(nu)=xF3(nubar), which essentially corrects
        the strange sea to zero), isoscalar
        correction, the Bardin radiative correction, the
        slow-rescaling correction, and the propagator correction
        (due to the mass of the W-boson).  The "-no-corr" files
        represents the effect of removing all five model corrections.

        These files are provided for the benefit of the people who
        have requested our structure functions extracted without a
        given correction (for example, we have received requests for
        our SF extracted without the isoscalar correction). 

        IMPORTANT: The model-correction files are different from the
        systematic-error files.  The systematic error files represent
        the shift in the structure functions due to a one sigma
        variation in the error of a quantity.  The correction files
        represent the shift in the structure functions due to 
        entirely removing one portion of our physical model.
        See W. Seligman Thesis, Appendix D, for more information.