When conducting an online survey, you will always go through the same stages, regardless of your actual questionnaire. There is a best practice stage guide to creating a survey. These stages are programming and testing the survey, collecting the data and then analysing that data. This page is dedicated to these four survey project stages. It will show you what to pay special attention to, and all necessary actions when moving from one stage to the next.
The first stage in the survey project process is the initial programming of the online questionnaire. Create a new survey and fill it with all of your required questions and other elements.
Another section feature is to randomise sections, pages or questions (if you have only one question per page).
While you are programming your survey, you should not activate it. The preview (either using the preview icon or using the general survey link) also works for non-activated surveys.
Your schedule for creating the survey should include plenty of time for the testing stage. In many cases, testing takes just as long as the programming of the initial survey, and is an important best practice step when creating a survey.
The purpose of the testing stage is to find any existing problems – in the questions, as well as in the survey routing – and then resolve them. Finding and resolving these problems at a later stage will always result in more work!
A recommended approach is to divide the testing into two steps. The first step would be to check the completeness and correctness (wording and question types) of the questions. When this is complete, the second step would be to check the routing – which means checking the filters, randomisations, placeholders, URL forwardings etc.
For this purpose, it’s useful to press the back button in the survey and change the answers, instead of starting each test case from the beginning of the survey.
In general, you can test the programmed survey both inside and outside of Survalyzer. When you are logged into your Survalyzer Account, you should preferably use the preview icon.
Testers who have no access (and should not have access) can use the general survey link instead, which you can find on the Distribution page. Make sure that the survey can be accessed more than once by the testers, but not by real participants.
If you want to conduct analysis using rawdata files, you should also check the generated rawdata after the first test interviews have been completed. As long as your survey is not activated, you can easily adapt variable names and codes to prepare analysis.
When you test your online questionnaire you should not activate it. The preview (either using the preview icon or using the general survey link) also works for non-activated surveys.
To begin data collection, participants have to be invited. For every online survey you can use the anonymous link (the same for all participants), or you can use individual / personal links (one link for each participant). A combination of both (some participants use the anonymous link, others use individual links) is also possible.
However, if you want to use individual links, you have to firstly create a sample. Creating a sample can be done during the programming or testing stage. More details about link distribution can be found on the page “How to invite others“.
Before you begin data collection, the testing stage should be complete. After completing the testing stage successfully, you can get your survey ready by activating it. When you activate a survey, all test data will be deleted and any quota counters will revert to zero. You should also check the settings for the back button, multiple participation, end date and authentication with token code. When this is done, you can start inviting participants.
Please note that it is essential that your survey is activated before participants are invited. In particular, invitations with individual links (including link lists) must wait until the survey is activated, as during the activation process already existing individual links become invalid.
Response Statistics and Online Reporting
On the page which lists our surveys, you can see all activated surveys by the status “On” (with a green background colour). To the right of the status information, you will see a number which tells you how many currently complete interviews there are. If you click on this number, you will go directly to the Analyse page of the survey. There, you can see the online report of all collected data so far. Additionally, you can download Excel report tables, PowerPoint charts and a PDF report or the raw data (either as an XLSX file or as SAV file for SPSS). More details about reporting options can be found on the page “Analyse“.
On the Distribute page you can click on each distributor in the list on the left hand side (e.g. an invitation mailing) to see the response statistics for this specific distributor.
It’s possible to give external viewers (e.g. clients) access to online reporting. To obtain login credentials just click on the “Share” button.
If, during the testing stage, not all errors are found and corrected (meaning it is necessary to make corrections post-activation), you can switch the survey to revise mode. To do this, click on the “Revise Survey” button in the top right corner (where you previously triggered the survey activation process). More Information about Versioning can be found under the topic “Survey Status“.
Once the data collection stage has finished, you should close your survey. You can either click on the “Close Survey” button or you can set an end date that has already passed. Closing the survey makes sure that no new data is collected and you can use the final data records for analysis. Details about reporting and file download options can be found on the Analyse page.
Note: Once data collection has finished, no questionnaire data will be deleted, unless you purposefully delete it.
You can analyse your data at any time. If you want to remove the collected data from your workspace, you can delete the whole survey or connected Panels (or single Panel members or even single Panel member interviews).
Follow these best practice recommendations to create a great survey.