System Suitability Criteria
System suitability criteria available within QuBAS for the 4PL model are:
Controls
-
Upper and lower limits can be set for the mean or geometric mean of:
-
-
By groups if present e.g. high, low
Reference standard
Goodness of fit of the model
(Automatic selection)
-
If there is more than one replicate at each dose: lack of fit error can be compared with pure error via an F test.
-
The level (critical p value) can be chosen by the user.
-
-
A 5PL model can be fitted to the data and compared with the 4PL fit via a likelihood ratio test (chi-squared test).
-
The level (critical p value) can be chosen by the user.
-
Parameter values
-
-
-
-
The confidence level for the each interval can be chosen by the user (no default setting).
Statistical significance of Dose response 
-
via a one-sample t test comparing with zero.
-
The level (critical p value) can be chosen by the user
-
Range
-
Range (difference between asymptotes): limits for the value of |
can be set.
-
Range (difference between asymptotes): equivalence limits for | 
can be set:
-
Range for the confidence intervals for | (no default setting).
-
The confidence level for the each interval can be chosen by the user (no default setting).
Variability of data
-
An upper limit for the variance, σ2Ref, mean square error (MSE) can be set.
-
An upper limit for σRef the root mean square error (RMSE) can be set.
Quality Control
Goodness of fit of the model
-
If there is more than one replicate at each dose: lack of fit error can be compared with pure error via an F test.
-
The level (critical p value) can be chosen by the user.
-
-
A 5PL model can be fitted to the data and compared with the 4PL via a likelihood ratio test.
-
The level (critical p value) can be chosen by the user.
-
Statistical significance of 
a) Via a one-sample t test comparing 
with zero.
-
The level (critical p value) can be chosen by the user
-
Parallelism with reference standard
a) Ranges for combinations of QC and reference parameters can be chosen.
-
Difference between equivalent parameters:
-
(
-
(
-
(
-
Normalised difference between asymptotes
-
-
b) Equivalence tests for combinations of QC and reference parameters.
-
Ranges for the confidence intervals (no default setting) for:
-
(
-
(
-
(
-
-
-
The confidence level for the each interval can be chosen by the user (no default setting).
c) Significance tests for lack of parallelism can be used:
-
-
The level (critical p value) can be chosen by the user.
-
-
-
The level (critical p value) can be chosen by the user (no default setting).
Relative potency of the QC sample
, and its confidence interval. (For calculation of these parameters, see 3.2.5 below.)
, 
, 
can be set on the log.
, 
, 
with the parallel model.