System level analysis determines intervals for Manufacturer/Models when they are placed in designated systems. The results are displayed in the System Evaluation Report, which is accessed by selecting the System Evaluation Report option from the Reports menu on the Main Screen.  Brief descriptions of the System Evaluation Report features and functions are given below. 

Scope Options

The scope of a System Evaluation relates to the Manufacturer/Models that will be sampled in performing the evaluation. The evaluation may be based on all Manufacturer/Models, only Models associated with the current Selections and Filters, or a specified range of Models.  If you elect the latter, a set of Selection Criteria are displayed.

Exclude Options
If it is desired to assess the impact of Class Level intervals on system performance, evaluations may be run with Class Level analyses alternately included and excluded.  It is also possible to assess the impact of Engineering Overrides on system performance by alternately including and excluding Engineering Overrides. 
 
Report Fields
The fields of the System Evaluation Report are: Current Interval -the currently approved interval. Assignment Level - the basis for interval assignment. The keys are C - Class, S - Subclass, R - Similar Equipment Group, B - Subgroup, M - Model Number, Y - System. Additional letter designators are D - Dog, G - Gem, E - Engineering Override, A - Interval Based on Method A3.  Number Calibrated - the number of service actions observed at or near the Current Interval.  Observed Reliability - the observed percent in-tolerance at or near the Current Interval.  Reliability Target - the reliability target for a given Manufacturer/Model. Test Results - whether IntervalMAX has passed or failed for a given Manufacturer/Model.

Running an Evaluation
The System Evaluation is run by entering a Confidence Level for the tests to be performed in the Enter Conf Limit box and clicking the Run Test button. The Confidence Level indicates the level of confidence in rejecting each computed interval. For example, if a Confidence Level of 0.9 is selected, intervals will not be "failed" unless there is a 90% confidence for doing so.

Test Results
The displayed test results are: Number Tested - the number of Manufacturer/Models with sufficient service history data at their respective Current Intervals. Number Passed - the number of Manufacturer/Models of the Numbers Tested that passed the tests at the Current Intervals.  Percent Passed - Number Passed / Number Tested multiplied by 100.  Average Reliability - the average observed in-tolerance percentage at the Current Intervals.  Standard Deviation - the standard deviation of in-tolerance percentages at the Current Intervals.  Default Reliability - The overall Default Reliability Target.