Design an experiment, fit a response-surface model, and build the ICH Q8(R2) design space — the multivariate region where every response meets its acceptance criterion — with a covariance-propagating Monte-Carlo edge of failure. Free, in your browser. Decision support for a qualified reviewer; not a substitute for formal statistical review or process validation.
1Factors & ranges (CPPs)
Factor name
Low
High
Unit
Add 2+ process parameters you want to study, each with a low and high setting. These become the columns of your design.
2Choose & generate the design
3Responses / Critical Quality Attributes
Response name
Linked CQA
Criticality
Lower limit
Upper limit
Target
Each response is a measured outcome tied to a Critical Quality Attribute. Set a lower and/or upper acceptance limit — the design space is where all responses stay within limits. Criticality drives the pCPP flagging.
4Enter measured responses
Generate a design (step 2) and define at least one response (step 3) to build the data-entry table.
5Model & design space
The two chosen factors are swept to draw the design space; any other factors are held at their center point. The edge of failure propagates BOTH model-parameter covariance and residual error — it correctly widens near the design boundary.
Enter data, then analyze.
This tool applies classical DoE modelling (OLS response-surface analysis, ANOVA, lack-of-fit and standardized-effect Pareto) and maps the ICH Q8(R2) design space as the region where all responses meet their pre-defined criteria, with a covariance-propagating Monte-Carlo edge-of-failure overlay. Results support, but do not replace, expert statistical review and formal process validation. The proven acceptable range and control strategy remain the sponsor's responsibility. Generated via the MolWard Platform.
