This page provides information and answers to frequent questions we receive from users.
- Can I substantively change an indicator?
- Should I add unknown as a disaggregation category?
- Can I add or edit disaggregation categories?
- Why do I have results for unassigned geographies?
No. Any substantive change to an indicator's definition will require creating a new indicator. Substantive changes include any revisions to:
- Data Source
- Indicator Type
- Geographic Disaggregation
- "Results are reported separately for each activity"
This is up to you and the way your organization collects data. In cases where you have several teams reporting on a single indicator despite not sharing the same disaggregations, adding an "unknown" category would help capture detailed information where it's available.
E.G: Every project reports on # people trained, but not everyone captures the information disaggregated by gender.
Unlike deleting disaggregations or disaggregation categories (which will delete associated data), DevResults will allow you to add new categories to an existing disaggregation, as well as edit the name of an existing disaggregation category. However, care must be taken to avoid changing the nature of the indicator or the definition of the data. Whether a change is methodologically valid depends on how you've defined your categories in documentation and in practice.
For instance, if you had a disaggregation for age with categories named '<18' and '18+', you could change the names to 'Youth' and 'Adults' respectively, assuming that you've clearly defined 18 years of age as the threshold of adulthood. All of the data previously associated with the '18+' category would now be associated with the 'Adults' category.
Similarly, if you disaggregate your training participants by the type of trainings you offer — e.g. 'finance', 'administration', 'advocacy' — and you intend to begin offering a completely new training — 'data collection' — there's no risk that any of the previous training participants would have ever completed this newly available training, and so your data integrity is maintained.
However, if you disaggregate "Household size" by the categories "1, 2, 3, 4+" and decide that going forward you want to add more granularity by using additional categories such as "1, 2, 3, 4, 5, 6+," you would be making a fundamental change to the original '4+' category. DevResults will allow you to make this change, but you must either re-enter data for all '4+' households to ensure accuracy, or create a new indicator with a new disaggregation to collect more granular data going forward.
Why do I have indicator results for an activity in geographies that are not assigned to that activity?
The answer to this question depends on the data source of the indicator you're referring to:
Direct entry indicators
To enter results in the first place for a direct entry indicator, a geography must be assigned to the relevant activity in order to insert the appropriate row into the data entry table. However, removing the assigned geography from the activity later on does not delete previously entered results; this is intended behavior.
Often, geographies will be added or removed to the scope of implementation as an activity goes forward. In most cases, users that remove geography assignments do not want to delete historical data, but only to remove rows from a reporting template to avoid confusion going forward. If they did want to delete results from a removed or unassigned geography, they could do so at any time by:
- (Re)assigning the geography
- Navigating to the appropriate reporting period(s) and deleting the relevant results
- (Re)unassigning the geography
Data table indicators
As noted in the article on Assigning a Geography to an Activity, data tables allow for users to make implicit geography-activity assignments simply by adding geographies to an activity-linked data table. In other words, if you add records to a data table with both an activity column and a geography column, you do not have to add the same geography to the activity directly for a second time. This prevents double data linking, which tends to be more onerous with data tables given the granularity of per-observation records.
This means that once again, results may appear, expectedly or unexpectedly, in geographies that are not explicitly linked to an activity. If this is undesirable, you can either remove the activity from the indicator, the record(s) from the data table, or filter out the errant geographies from whatever visual or report you are creating.
Formula indicators follow their component indicators, so no special considerations are necessary. If they are comprised of direct entry component indicators, explicit activity-geography assignments — whether active or historic — will govern the presence of results data. If they are comprised of data table component indicators, implicit geography assignments will govern the presence of results data. If both types of component indicators are present, both situations may apply.
Didn't answer your question? Please email us at email@example.com.