Validating Evidence

The evidence framework provides support for adding validations to the insert evidence, modify evidence, and apply evidence changes sub-patterns for each evidence type. These sub-patterns access the informational manager which allows for standard warning and error validations. Warnings should be used to inform users of important information that must be updated or fixed at some point in the future before the evidence is applied. Errors are used when a process should not be completed until the validations are executed successfully.

The use of warnings should be considered when adding validations for the insert and modify evidence sub-patterns, in particular, for evidence relationship validations. This will allow the user to focus on the capture of evidence requirements without being forced to follow certain restricted paths in maintaining related evidence. All validations for the apply evidence changes sub-pattern must throw errors rather than warnings since the activation and removal of evidence can impact case eligibility and determination.

One obvious benefit to validating evidence is the reduction in errors. Validations can also be used to check for gaps in the evidence and to look for duplications. For example, there may be certain evidence types which require one active evidence record at all times such as an active primary address. Validation can be used to make sure there is no gap, i.e., that there is always an active address record. To look for duplications, an additional validation can be implemented which does not allow more than one evidence record, e.g., the primary address evidence record, to be active at the same time.

As part of designing an evidence solution, it is important to recognize the strong relationship between validations and how the rules loaders should be designed to load the evidence data. For example, consider a validation which states that a person cannot collect income more than once from the same employer for the same time period. Given this validation, the loader only needs to search for one piece of income data at a point in time. On the other hand, if the validation stated that a person could collect multiple salaries from the same employer over the same period of time, the rules loader needs to look for a list of items to load.