For users that perform additional validations outside of Survey Solutions, a frequent request is a compilation of validations in Survey Solutions.
While this compilation can currently be created manually by copying from Designer and pasting into some validation repository (e.g., Google Sheets), it would be convenient to have in Designer an automatically populated table with the following columns:
- Question. That is, the variable name of the question where the validation appears. (For validations that appear on static text, perhaps show the first N characters of the text or its questionnaire location.)
- Error or warning message. That is, the message text associated with the validation. Absent a validation “label”, this is the best descriptor of a validation.
If possible, other columns could prove useful for filtering or knowing more about a table entry without navigating to its occurrence within the questionnaire:
- Error/warning. This would communicate the validation property: error or warning.
- Section. This could be a column added by the user if needed. This could be a useful filter to see which validations exist in a given section.
- Validation condition. In Designer, this would be helpful answering the question of which questions are involved in the validation. As externally saved documentation, this could more fully explain each validation.
Upon clicking on an entry in validation compilation table table, the Designer user would be taken to where the selected validation appears.
Since this is an advanced feature, perhaps it could be located alongside other advanced features in the tool bar on the left-hand side of the Designer screen.
To make this compilation as widely useful as possible, it should be able to be exported to a format that is readily usable by survey team members who do not use Designer.
Having such a compilation would be useful in several ways:
- Reminding CAPI qnr designers which validations have (and have not) been implemented
- Informing survey managers, who manage CAPI qnr developers, of the same
- Informing data analyst, tasked with programming additional validations in some statistical software package (e.g., Stata, R, etc.), which validations Survey Solutions has already covered.
- Documenting in consolidated form which validations are part of the questionnaire
While I hear this a lot in my own work, my sample may be biased. Would this be useful to others?