1. Determine research method
The PIEL app presents two options for running surveys.
(a) Scheduled Survey
A "Scheduled Survey" presents the survey to the participant at various times over the course of a day or week. With this option the participant does NOT initiate the timing of the survey.
A scheduled survey is created by including the following line in the control file.
(b) Run-once Survey
A "Run-once Survey" is initiated by the participant. It can be started multiple times but each time will only run once.
Run-once is the default mode of the PIEL Survey app. Alternatively, you can set it run-once explicitly using the following line in the control file.
2. Decide survey distribution method
In most cases, the control file will be distributed using email. It can also be directly loaded onto a device using messaging apps that allow attachments. File sharing using Airdrop or Bluetooth can be useful if the researcher provides participants with mobile devices for the survey. For larger studies, you may also opt for a web server distribution method.
These options are detailed here.
The best participant experience is to include a "subject-id" in each file. This requires the researcher to make a small change and add a subject ID line to each individual participant's control file. An example of how to set the ID follows.
If a participant's subject ID is not included in the control file, he or she will be asked for their ID when starting the survey. The researcher will need to inform participants of what ID to use prior to starting the survey.
Note that for a scheduled survey, the ID only needs to be entered when the series of surveys is first started. It is saved for subsequent survey runs.
3. Decide data collection method
In most cases, researchers collect data using email but file sharing can also be used, typically when a researcher provides participants with devices. It is important that participants have a valid email account on their device if the email method is used.
A participant can email results each time a survey is completed for a run-once study, or else at the end of the study for a scheduled survey. Results are progressively added to the data file so the last data file will contain all previous data. In the case of a run-once survey, if a participant ID is entered manually, a new file is created for each participant ID entered.
4. Prepare questions
This is perhaps the most critical part of your survey research preparation. The question type, wording and order of the questions and responses are important decisions. If you decide to use branching logic, it is essential that you create the correct logic so the participant is not presented with irrelevant or incorrect questions. For scheduled surveys, you may also decide to limit the amount of time a participant has to open (max-delay) and/or to complete the survey (max-duration).
The PIEL Survey app helps your survey design in the following ways:
- There is only one question allowed per screen. This ensures a participant's focus on each question.
- The participant must interact with each question before they can navigate to the next question. Questions cannot be skipped. If you wish to allow a "no response" option to list or checkbox questions, you must add this as a response option for this particular question.
- Large fonts are used to allow participants to easily answer questions during their daily activities, whether on a bus or in a meeting. Questions and responses should not be too long.
- Progress between questions is rapid with the enabling of swiping between questions. This encourages participants to finish the survey.
The quality of results obtained from surveys is very dependent on the questions asked. Further discussion on the design of questions in the PIEL Survey app can be found here.
5. Develop survey control file
Starting with a blank canvas can be difficult so we recommend that you develop your first survey file from a pre-existing survey. Make sure that the file is saved as a text file with UTF-8 encoding (usually the default). The file must be saved with a name that has a ".survey" extension. Please be aware that some software tries to add a ".txt" extension. If so, rename it with a ".survey" extension.
See the following links for detailed instructions on how to create a control file.
You may also find an easy starting point is to follow this video tutorial.
6. Researcher testing
We recommend that you thoroughly test your survey prior to sending it to participants. The Control File allows a lot of flexibility in how you structure and run a survey, so it is important that you ensure it collects the data you want in the way that you want.
In order to facilitate testing, it is useful to enable test mode. This is done using the following entry in the control file.
After testing, you may choose to disable testing before starting the actual research study. This will remove the "Test" button from the user interface.
7. Create instructions for participants
The PIEL Survey app is very intuitive for a participant and an experienced mobile device user will only require minimal instructions. However, we advise that you provide instructions to participants covering the following issues:
- Download the PIEL Survey app
- Import and select survey
- Preparation of device
- When to start survey
- Email results
If you are providing devices and intend to provide a device with the file already imported, some items in the above list will not be needed.
As a starting point, we have provided an example participant information sheet.
8. Secure approval
Some research projects may require approval of your institution's Ethics Committee or Institutional Review Board. Numerous research projects using the PIEL Survey app have been through this stage at well known universities and have been published in peer reviewed journals. To assist in this process, please see our confidentiality information.
9. Conduct pilot study
It is highly recommended that you test your survey questions, survey behaviour and participant instructions in a pilot study. This may alert you to problems that you had not thought of when developing the survey.
10. Conduct research study
You are finally ready to start! Key steps to commencing the project are:
- If you are inserting "subject-id" in the survey file, you will need to individually edit the control file for each participant.
- Email the control file and instructions to participants.
- Maintain contact with participants to make sure everything is progressing smoothly.