The data import logic also supports a process where existing respondent data can be updated afterwards by making an additional data import, containing the updates only. In cases where you need to update incorrect data in a project with many data sets: Instead of importing all data files again, which is time consuming; one data file that only includes the concerned respondents is imported. For example, the originally imported data set looks like the first data table below. As shown, the Segment variable is missing data in a few rows. A new data set is imported which includes the missing data as shown in the second data table below. During the data import, the second data set is defined as “Update to existing data set(s),” so the second data set will not act as a new data set, it will only update the existing data sets in the project. The third data table below shows the result of these two imported data sets after data activation.
Note: Update can be used for only adding variables or only changing case data too (it doesn't have to happen both at the same time).
In the example above, the Respondent ID variable has been used as the mapping key during the mapping process between data sets 1 and 2. The variable to use is flexible, it does not need to be the Respondent ID; it can be any variable in the project. It does not need to be a unique variable either. For example, if importing a file as “Updating existing data set(s)” with two columns (like in the example below) and using the “Team” variable for the mapping between the data sets, all respondents in “Team 1” will get the value 22 in the Index A variable.
1. Select the Import tab and in the dropdown menu select the file type and press
Browse to search for the file. Select the file you want to import as update to existing.
2. On the data import page below the Batch type, select Update to existing data set(s)
3. Select a variable to use as the mapping key.
If the data set is a new, regular data set, the check-box should not be ticked. In the data activation page, it is shown if the data set is an “Updates to existing data set(s).” These data sets are active in the same way as other data sets. The image below shows the data activation page.
The updates are made during Data Activation. Both the ordinary data set and the date set including the updates are necessary to be activated.
- A data set defined as “Updates to existing data set(s)” can only update "regular" data batches imported prior the data set with updates was imported. Regular Data batches imported after t will not be affected. Once you select "New data set and updates to existing data sets", it can contain both new data and updated data.
- One data set defined as “Updates to existing data set(s)” can contain information that affects multiple data sets prior to being imported. It is not a 1-on-1 relationship between originally imported data sets and those with updates.
- One row in an “Updates to existing data set(s)” can update multiple respondents in the prior imported data sets if the mapping key matches the multiple respondents.
- Only the variables present in the “Updates to existing data set(s)” will be updated during data activation. E.g. If the original data set contains 20 variables, and the “Updates to existing data set(s)” only contains 2 variables, only these 2 variables are updated. The other 18 variables are not cleared/updated.
- An “Updates to existing data set(s)” can include metadata updates. This is the same logic that is used when importing a data set with new data.
- If a RespondentID, exists in the “Updates to existing data set(s),” but not in any original imported data set, the data cannot be activated.
- If you need to update one answer of a multiple, you will need to add all columns which define the multiple in Dapresy. So in case you want to add answer 4 to your multiple, you need to have 4 columns in your file. Otherwise it will overwrite the multiple and only include answer 4.
- Updating a "Date" variable with system missings/blanks is not possible.
Note: using "Updates to existing functionality will slow down your data activation process! If it's easy to replace complete files, this is definitely a recommended solution when taking performance into account.