Using Reporting Data in Reportal

In Reportal, if you have a report that has only one data source and was using BitStream files, you can change this to use Reporting Data instead. To do this, in Reportal, go to the Data Source toolbox, right-click on the root folder and select Switch Data Source to Hub. Note that this is a one-way operation - once you have changed the data source to use the hub it cannot be changed back to use BitStream files. Also note that this will apply to all reports using the same data source. If you want to switch a live report to “Reporting Data”, it is recommended to duplicate the report first, and switch the duplicate to “Reporting Data” and verify that the report works as it should, before switching the original.

Benefits of using SmartHub/Reporting Data and Differences Compared to BitStream files

If selecting to run Reportal reports on “Reporting Data” or a hub set up in SmartHub, the following benefits can be achieved:

• Nearly real-time data updates through continuous synchronization of data from survey databases etc. that is triggered every 5 minutes if there are changes in the datasets.
• Improved performance and scalability, especially on large datasets.
• Consistency between data displayed in hitlists and aggregated tables. When using BitStream files, hitlists will get data directly from the survey database, which may not have been written to the BitStream files (yet). With SmartHub/Reporting Data, hitlists and aggregated tables both are querying the same back-end, so there is consistency in results reported.
• New analytics functionality, like ability to set aggregation levels on loop data/one-to-many relations (“hierarchical data”).
• With reports based on a hub configured in SmartHub: greatly enhanced capabilities for combining and mapping different data sources.

There are some key differences between reporting on SmartHub/Reporting Data and BitStream files to be aware of:

• SmartHub/”Reporting” Data only support the “extended” table engine. If you have an existing report on table engine “version 1”, you need to switch it to the extended table engine first, and then switch to Reporting Data. It is recommended to make a duplicate of the report, and switch that to the extended engine and verify first, before switching a live report to the extended engine.
• Reports with multiple projects in the source are not supported on “Reporting Data”. Base the report on a hub configured in SmartHub instead.
• It is not supported to set up Joins and Unions in the data source inside Reportal for reports on “Reporting Data”/SmartHub. Instead you can set up a hub in SmartHub with a “Combined data source” (union) or linking surveys to a Contact database (join).
• Timezone Offset is not supported on SmartHub/”Reporting Data”.
• Note that BitStream and “Reporting Data”/SmartHub will do data synchronization at different points of time. BitStream files are generated through scheduled data updates, whereas SmartHub by default will synchronize data continuously when there is changes/new data in the survey db. So for a live survey, data may differ because the synchronization has been run at different points of time for BitStream and “Reporting Data”/SmartHub. So to compare similar datasets, you can for example apply a global filter on the report that uses “Reporting Data”/”SmartHub” as a date filter based on time less than the time of the last BitStream synchronization.
• “Reporting Data” and SmartHub will bring in all records from the survey database. BitStream files may be filtered on status, and do not include “screened” responses. So to achieve the same filtering, you will have to set a status filter (for example status = ”complete”) on the report level in Reportal after switching to use “Reporting Data” or a hub.
• Deleted categories: If you have data stored in the survey database for categories (answers) that have later been deleted/removed from the questionnaire, BitStream will be including those deleted values when counting total number of records. When reporting on SmartHub/”Reporting Data”, the deleted values will be treated as no response, and not be included in the total.
• Filters will not affect the level of aggregation used in a table when report is based on “Reporting Data”/SmartHub. The default rule used for aggregation level is that the default aggregation level will be the lowest level, so if a variable inside a loop is added, the aggregation level will automatically be set to the loop level. For a report using BitStream this also happens when applying a filter that references a loop variable. For reports based on SmartHub/Reporting Data, the filter will not affect aggregation level.

If there are inconsistencies between values in the survey database for background variables in respondent data and response data, for example due to survey scripts, API updates, updates in data editors or Data Processing/data imports on one table, but not the other, there will be differences in results produced by BitStream and SmartHub/”Reporting Data” if you set a table to “Use Respondent Data”. This is because SmartHub/”Reporting Data” will be fetching values from respondent data for background variables when you set “Use Respondent Data”, whereas BitStream is only fetching values from respondent data for “Not answered” respondents, that don’t have any response data, and will combine that with response data for respondents that have answered the survey (partially or complete).