Making changes to published surveys

We recommend that before publication surveys are thoroughly tested using the Preview option.

It is sometimes useful to provide colleagues with the survey's Preview URL to allow them to also test the survey.

The Preview will allow the main part of the survey to be reviewed, i.e. this will allow you to check for spelling and typing errors, check that questions are not ambiguous and that the most appropriate answer formats have been used. (The most common mistakes are using single radio instead of multiple check boxes and vice versa).

The branching logic or mandatory field status of a survey cannot be tested until the survey is published, but these are both survey characteristics that can be changed while a survey is live.

When publishing a survey we would also recommend that an extra few days are added to the publication period as if changes are necessary this is often more cost effective than extending the publication period.

It is highly recommended that a survey is tested again once published. Providing there are twenty or less responses the test data can be purged using the Purge facility before live data is recorded.

If a published survey needs to be changed you can make some changes that do not affect the data structure as often as you like while the survey remains published; these are limited to the survey's characteristics such as style, branching logic and mandatory status.

To make changes to the characteristics simple use the Compose option and then choose the Compose Characteristics option.

Survey Galaxy also provides a 3 day grace period where changes that could affect the data structure can be made to published surveys without charge, although 1 day of publication is deducted for each re-publication.

To make changes that could affect the data structure use the Compose/Revise option.

When a published survey is changed via Compose/Revised the survey will not be available to respondents until the survey is re-published. To ensure integrity a new data-set will be created to ensure that there is no data miss-match. All previous data-sets will be saved and will remain accessible online.

If email invitations have been sent out containing a URL to the survey and the published survey is later composed/revised and republished the URL will always point to the latest publication of the survey.