What?
A readyforwhatsnext modelling project undertakes novel analysis of youth mental health topics by using, enhancing and/or authoring model modules, datasets and executables. Each ready4 modelling project has its own unique funder(s), governance, objectives and team. The links between modelling projects are in the form of a common framework.
Undertaking modelling projects will help us achieve our following priority goals:
-
Priority 3. Applying readyforwhatsnext.
-
Priority 5. Extending readyforwhatsnext.
Who?
Modelling projects should typically be led by a researcher who may or may not be a modeller. The core project team will always include modelling expertise and, should authorship of new modules (or extensions to existing modules) be required, will also need to include coders. Advisory structures to engage community members and planners are also recommended.
How?
There are three main steps in implementing a ready4 modelling project.
Step 1: Develop model
Each project’s computational modelis constructed by adopting one or more of the following strategies:
- selecting a subset of existing readyforwhatsnext modules and using them in unmodified form;
- selecting a subset of existing readyforwhatsnext modules and contributing code edits to these modules to add desired functionality;
- selecting a subset of existing readyforwhatsnext modules and using them as templates from which to author new inheriting modules (which can be code contributions to an existing module library or distributed as part of a new library; and/or
- authoring new ready4 modules (most likely to be distributed in new code libraries).
As part of the validation and verification process for all new and derived modules, tests should be defined, bundled as part of the relevant module libraries and rerun every time these libraries are edited.
Step 2: Add data
By data we typically mean digitally stored information, principally relating to model parameter values, that can be added to the ready4 computational model to tailor it to a specific decision context (e.g. a particular population / jurisdiction / service / intervention) and set of underpinning beliefs (e.g. preferred evidence sources). Data for a ready4 modelling project will be from one or both of the following options:
- finding and using existing open access data from other ready4 projects;
- supplying new project specific data, appropriately describing these data and (for non-confidential records) sharing these data publicly.
Step 3: Run analyses
ready4 project analyses apply algorithms contained in ready4 modules to supplied data to generate insight and can be implemented by:
- adapting existing replication programs;
- authoring new analysis programs; and / or
- developing a user-interface to allow non-technical users to run custom analyses.
When reporting analyses, using a reporting template can be useful.