[ sedfit help home ]

[ Data ] [ Copy ] [ Display ] [ Model ] [ Parameters ] [ Run ] [ Fit ] [ Generate ] [ Statistics ] [ Notes ] [ Options ] [ Help ]

## the statistics menu

Sedfit has a few functions
for statistical analysis of the results from modeling the data.

Besides the Runs Test,
which is calculated for every fit, Sedfit
has functions for calculating confidence intervals for sedimentation
coefficients and molar mass values. This is done by the method of
constraining the parameter of interest to sub-optimal values, fitting for all
other parameters, and monitoring the increase of the rms error of the fit.
The cut-off increase is calculated using F-statistics.
While this is process is automated for s and M in the non-interacting
discrete species model, only partially automated functions are available for
mapping two-dimensional error contours of monomer-dimer
self-association models.

For more complex problems or confidence levels for parameters in other
models, they can be calculated by moving the parameter of interest manually to
sequentially less optimal values, until the cutoff value of the rmsd increase is
reached (which can be obtained from the F-statistics
calculator function). This strategy is explained in detail by Johnson
& Straume (ref1).

A different way for analyzing the statistical properties of the results is
recommended for the size-distribution analyses, which is Monte-Carlo
simulations.

Confidence Intervals

Error Contours

Monte Carlo Simulations

related topics: F-statistics, Runs
Test

**References:**

(1) M.L. Johnson and
M. Straume (1994). Comments on the analysis of sedimentation equilibrium
experiments. In: Modern Analytical Ultracentrifugation (Schuster TM and
Laue TM, eds) Birkhauser, Boston, p.37-65