7
Statistical parameters
The closeness of fit between the experimental data and the fitted curve is formally
described by a set of statistical parameters, described in the table to the right.
Parameter Description
Chi-square A measure of the closeness of fit, calculated as the average squared residual (the difference between the
experimental data and the fitted curve).
where r
f
is the fitted value at a given point
r
x
is the experimental value at the same point
n is the number of data points
p is the number of fitted parameters
chi-square =
n – p
( r
f
– r
x
)
2
n
∑
1
Standard error (SE) A measure of the parameter significance, reported separately for each parameter.
The parameter significance indicates the extent to which a change in the parameter value affects the closeness
of fit. A parameter with low significance can have a wide range of values without affecting the fit.
T-value The parameter value divided by the standard error. This can make it easier to compare the significance of
parameters with widely differing absolute values.
Uniqueness (U-value) An estimate of the uniqueness of the calculated values for rate constants and R
max
. For correlated parameters,
the fitting procedure can determine their relative magnitudes but not absolute values. For example, knowing
the affinity gives the ratio but not the values for rate constants. The U-value is determined by testing the
dependence of the fit on correlated variations in pairs of parameters, and is reported as a single value for the
whole fitting. U-values above about 25 indicate that two or more of the parameters (rate constants and R
max
) are
correlated and the absolute values cannot be determined. If the U-value is below about 15 the parameter values
are not significantly correlated.
Note: Some Biacore systems do not report a U-value.