Importing data from another survey
You can import and export data, variables and codes between surveys even if the variables and codes are not identical. The easiest way to do this is using a database link to import or export the data.
This method can be used to merge surveys:
- As a precaution, archive your survey first before merging the data.
- From the Survey Overview window, open the survey that you want to add data to.
- On the main toolbar click Database Links to open the Database Links dialog.
- Click New to open the Database Linkage Wizard which guides you through the process of creating a new database link.
- In the Linkage Type section, select Import from Snap Survey and then click Next
- A list of all the surveys in the default directory is shown. Select the survey you want to import from and then click Next. (If the survey you require is not shown use the Browse button to find the appropriate directory.)
- On the next page of the wizard select the type of Import/Export option you want:
- Join cases where the link variables match
- Only import/export data where the value in the link variable matches between the two surveys (E.g. a survey containing customer ID numbers matches the information in the link survey with the same ID number.) Snap Desktop imports all requested data where two ID numbers are matched, and ignores all data in cases where the ID number does not match.
- Join matching cases and append unmatched cases
- This option works as above, and also adds the unmatched data as new cases at the end of the file.
- Append unmatched cases
- This option only imports/exports cases that do not find a match between the link variable in the surveys.
- Append all cases
- This option imports/exports every case of data into your survey as a new data case. It does not try to match the link variables.
- Select the link variable in your current survey marked as This Survey and then select the variable containing the matching data in the Link Survey you are importing from. The link variables must contain a value that is unique for each case, and which matches between the two surveys. For example they could both be ‘Customer ID’ variables. (Link variables are also known as database keys).
- Click Next to move to a list of variables in the Link Survey to import from and variables in This Survey on the right. Map the variables between the surveys. A default mapping between the variables is shown, where possible. To change a mapping, select a row in the right hand column, and click to display a list of available variables in the Link Survey. Select a variable to load data from, or select blank to load nothing.
If you match variables with multiple responses, the possible responses are shown below the variable list. Match the responses in the same way as you match the variables.
The example shows an extra code, Wine/Beer, is an extra code in the survey that you are importing data to. You can assign Wine/Beer to the Other code.
You could also create a new code in the current survey for Wine/Beer by right clicking the row and selecting New code from the context menu.
You can create a new variable to match an existing variable in the same way. Right-click in the right hand row, and select New variable from the context menu. Any variable that you create in this way can be modified or deleted.
- Click Next to display a summary of how the link will merge the data.
- Check the details are correct, and enter a Name for the database link.
- Click Finish to return to the Database Links dialog now showing the new link.
- If you wish to alter any part of the link, click Modify. To delete a link, click Delete. To make a copy of the link, click Clone, and then modify the new link as required.
- Select the link and click Run to import the data. A summary report is displayed. Click OK to close it.
- Now you can check your data by looking at a summary of the results. In the Questionnaire Design window, click Data View Mode and choose Counts on the toolbar. Scroll down through the questionnaire checking the figures for each question are correct.