Data Quality

Validation Rule Group A validation rule group provides a mechanism for classifying related data elements. Another advantate of using validation rule grops is that it can later be run separately, contrary to running all validation rules.

Validation Rule Analysis The validation rule analysis function will test validation rules against the data in the system. Validation violations will be reported in cases where the condition defined through the validation rule expression is not met, ie. the condition is false. First, enter a start date and an end date for which data should be included in the analysis. The date picker widget may be used to select dates. Second, choose between including all validation rules or a single group. Third, choose between including the selected organisation unit only or the selected organisation unit with all children in the analysis. Fourth, select the organisation unit. Finally, click validate. The analysis process while run for a while dependending on the amount of data that is being anaylysed. If there were no violations of the validation rules a message saying validation passed successfully is displayed. If there were validation violations they will be presented in a list. The organisation unit, period, left side description and value, operator, and right side value and description for each validation violation are displayed. The show details icon can be clicked in order to get more information about a validation violation. This will open a popup screen that provides information about the data elements included in the validation rules and their corresponding data values. This information can be used in order to correct incorrect data. The validation violations can be exported to a PDF document by clicking on the export to PDF button and to a Microsoft Excel workbook by clicking on the export to workbook button.

Std Dev Outlier Analysis The standard deviation based outlier analysis provides a mechanism for revealing values that are numerically distant from the rest of the data. Outliers can occur by chance but they often indicate either measurement error or a heavy-tailed distribution. In the former case one wishes to discard them while in the latter case one should be cautios in using tools or interpretations that assume a normal distribution. The analysis is based on the standard normal distribution. First, select the from and to date for the data to include in the analysis. Second, select the data set from which to pick data elements from. Third, select all or some of the data elements in the data set by double-clicking or marking them and clicking the add/remove buttons. Fourth, select the parent organisation unit to use. All children of the organisation unit will be included. Fifth, select the number of standard deviations. This refers to the number of standard devations the data is allowed to deviate from the mean before it is classified as an outlier. The possible outlier values discovered will be presented in a list after the analysis process is finished. The data element, organisation unit, period, minimum value, actual value, maximum value will be displayed for each outlier. The minimum and maximum value refers to the border values derived from the number of standard deviations selected for the analysis. Each outlier value can be modified directly in the analysis result page. The value can be modified by clicking inside the corresponding field in the value column, entering a value and then navigate away from that field either by clicking tab or anywhere outside the field. The system will provide an alert if the value is still outside the defined minimum and maximum values, but the value will saved in any case. The field will have a red background color if the value is outside the range, and a green if inside. Each outlier value can be marked for further follow-up by clicking the star icon.

Min-Max Outlier Analysis The min-max value based outlier analysis provides a mechanism for revealing values that are outside the defined minimum and maximum values. Minimum and maximum values can be custom defined or automatically defined by the system in the data entry module. See section about Std dev outlier analysis for further details on usage.

Gap Analysis The gap analysis provides a mechanism for revealing gaps in the data. A gap exists in the context of a data element and organisation unit. A gap is defined as a period with preceding and succeeding periods which have registered data values, but without registered data values itself. Such a gap might indicate a data capture error or omission and could be further investigated. See section about Std dev outlier analysis for further details on usage.

Follow-Up Analysis The follow-up analysis function will list all data values which are marked for follup-up. A data value can be marked for follow-up in the data entry module and in the other validation analysis variants in this module. See section about Std dev outlier analysis for further details on usage.