(For fundamental information regarding compliance scores, see AMP Compliance Scores)
With the Spring 2014 update to AMP we have introduced projected compliance scores, to remove the impression that reports that have not had manual testing performed are more compliant than those that have had manual testing performed. To do this we will find the average delta between the compliance score before and after manual testing for all reports that have had manual testing performed. We use this average to reduce the reported compliance level on audits that have only run the automatic tests.
To determine the Delta for any individual instance of AMP we look at each report that has had manual testing performed. For each report that has has manual testing done, we look at the compliance, and the automatic-only compliance on a per-standard basis and also an overall (across all standards) compliance. We then roll this up to the project, and then up to the organization. Finally, after all the organizations on the instance have been calculated, we average the delta's (per standard and overall) at the instance-level. This is then the delta we use to reduce the compliance scores when manual testing not been completed. As a result of this system the delta's are calculated on a per instance level in AMP (in the amp.ssbbartgroup.com instance this is handled on a per root organization level), so each customer's projected compliance delta is based only off of their results. This means that the delta is customized for each customer providing you with more accurate projections.
To keep the projected scores as accurate as possible the delta percentages are updated every 2 minutes, as new data becomes available. In AMP we also take into account partially completed manual testing when displaying your projected compliance score. When manual testing is only partially completed, the % of modules that have had manual testing completed is taken into account when applying the delta-reduction. For reports that have not had manual testing completed, we apply a reduction (the delta) to the compliance on a per-standard and to the overall report compliance. The amount of delta applied to each report is taken into account when looking at how much delta to apply to the project-level compliance. For example if you have a report with 10 modules and complete manual testing on one of those modules we apply only 90% of the delta since 10% of the manual testing has been completed.