The Analysis Parameters Screen is activated by selecting the Analysis Parameters option from the Controls Menu on the Main Screen.  The Analysis Parameter screen include tabs for setting/selecting Reliability Analysis Controls, Interval Criteria and Options, Outlier Criteria, Date and Time Constraints, Instrument Class and Group Development Criteria and Method A3 options.  Brief descriptions of the features and functions for each screen tab are given below. 

Analysis

The Analysis tab includes the various control variables that specify reliability analysis criteria. These include: Minimum Samples per Cell - The minimum number of resubmission times needed to define an observed in-tolerance probability.  Minimum Cells - The minimum number of cells needed to fit a reliability model to data. Minimum Model Resubmissions - The minimum number of cells required for analysis at the manufacturer/model level. Minimum Class Resubmissions - The minimum number of cells required for analysis at the class level. Time Series Outlier Test Confidence - The confidence level for rejecting short and long resubmission times within a reliability analysis time series (series of cells). Interval Change Confidence - The minimum confidence level required for rejecting the current interval in favor of a new one.

In addition to analysis criteria, there are two options: Remove Time Series Outliers - This option is selected if it is desired to remove short and long resubmission time outliers.  Assume Renewal at Every Service - If this is option is selected, then it is assumed that items are adjusted or otherwise renewed at every service, regardless of renewal codes entered by servicing technicians. 
 
Intervals
The Intervals tab is used to place Interval Assignment Constraints. These constraints may be enforced as the following control parameters: Max Allowed Interval - The maximum interval allowed by a using activity.  Min Allowed Interval - The minimum interval allowed by a using activity. These are administrative rather than a technical constraints.
 
Another set of control parameters are the Interval Change Criteria.  Interval change recommendations may be triggered whenever a computed interval differs sufficiently from the current assigned interval. An interval change is triggered by the any one of the following criteria: Fixed Number of Days - An interval change is implemented when a recommended interval differs from the current interval by some specified fixed number of days.  Percent of Current Interval - An interval change is implemented when a recommended interval differs from the current interval by some specified percentage of the current interval. Outside of Confidence Limits - An interval change may be implemented when the current interval falls outside the Analysis Confidence Limits when this box should be checked. 

The Reliability Model Mode options allow you to choose between using the standard reliability models incorporated into IntervalMAX or using intercept models.  An intercept model is one where the reliability may be less than 1.0 at the beginning of the interval.  Users with experience in reliability modeling can specify parameter constraints to be used instead of default values.  This is done via the Initialize Parameter Constraints Screen, which is accessed by clicking the Set Parameter Constraints button.  

In addition, the Interval Display Units can be selected from a drop-down list.  These are the units for displaying intervals and resubmission times. 

Outliers

The Outliers tab is used to set criteria for setting Outlier Test Confidence Levels for Dogs and Gems, Suspect Users, Suspect Labs, and Suspect Technicians.  There are also options for setting Dog/Gem Interval Multipliers to lengthen or shorten intervals assigned to Dogs and Gems.  Lastly, there are options for Dog/Gem Data Removal to prevent Dog and/or Gem data from skewing analysis results for mainstream items.  In particular, the occurrence of uncharacteristic out-of-tolerances can lead to a recommended interval that is substantially shorter than what is required for mainstream items within a grouping.

Dates

This tab used to set Resubmission Time Constraints for the following parameters: Earliest Service Date, Short Resub Time Limit and Long Resub Time Limit.

Classes

The option to implement automatic instrument subclassing is set using the Classes tab.  Once subclasses are formed, they are re-tested to see if any should be combined together. Confidence levels for creating subclasses and recombining them can be entered in the appropriate boxes.

Groups

The option to implement automatic subgrouping is set using the Groups tab. Once subgroups are formed, they are re-tested to see if any should be combined together. Confidence levels for creating subgroups and recombining them can be entered in the appropriate boxes.

Costs

The Costs tab of the Analysis Parameters screen provides a shortcut way to specify average hours per service, hourly rate and average materials cost per service. These are default values used by IntervalMAX in cases where specific cost data are not available in imported service history records. Cost data are used in projecting workload and cost changes due to interval changes.
Detailed cost data, by manufacturer/model, may be viewed and edited on the Cost Variables Screen accessed by selecting the Cost Variables option from the Controls menu on the Main Screen.
 

Method A3

This tab is used to specify one of the following Method A3 implementation options: Enable Method A3 - Placing a check mark in this box activates the Method A3 algorithm for an analysis run. Interval Change Confidence - This is the confidence level for rejecting the in-tolerance rate for an analysis candidate.  In general, the higher the reliability target, the lower the rejection confidence. Recommend when different from Last Approved - Method A3 intervals are recommended only if they represent a change in the status quo.  Recommend when different from Average Resubmission Time - This option is selected for cases where no Last Approved interval is available in IntervalMAX, or when it is desired to recommend an interval if it differs from the historical average.