Warning and error icons have been added to the editor to help identify specific questions that may be problematic.
In addition to an individual icon appearing to the right of each question, warning and error icons with counts appear at the top of the question list to provide a summary for the report.
A warning identifies a question as likely to have an issue or limitation with functionality.
The user is informed that additional action(s) may be required to correct the potential issue, but the question is still included in the dataset and can be used in Halo Reports.
Some examples of warnings are:
The question dictionary does not exist (not applicable to verbatim questions);
The dictionary exists, but is incomplete (not applicable to verbatims);
The same code exists in 2 dictionaries with a different label, stat value, ‘no answer’, or ‘missing’ designations.
An error renders the question unusable and prevents it from being included in the dataset.
However, imported data files that include questions with errors are still processed and converted into a Halo Reports dataset.
Any questions with errors will need to be corrected before they can be included in the dataset and used in Analyze.
Some causes of errors are:
A multichoice question includes variables/responses with inconsistent dictionaries;
An evaluation set question includes variables/responses with inconsistent dictionaries;
The question dictionary includes duplicate codes;
A question that is based is using an invalid definition.