Describe the data you need to collect
Before you start your project, you should consider where the data are likely to be stored. This is both your collected data (i.e. measurements or observations) as well as what metadata you will need to record (this is information about your project such as the start date, number of participants etc.). If you consider this early you can format your data appropriately from the beginning, saving you time (and preventing potential headaches!) when you want to upload your data to store it. If you can, always try to ensure that your data are stored in an appropriate repository. Please refer to the Storing and sharing your results section for more information.
Define which methods you will use to collect data
There are several ways data can be collected for a citizen science project. Ensure your chosen methods are accessible to participants, and that the collected data can be easily submitted. The image below shows the most common forms of data collection methods.
It is important to assess each method in terms of:
- Suitability and usability for your participants
- Web portals
- can provide easy access for participants to upload collected data and images, and work as a simple directory to promote the project, engage with participants and share results
- can also be an avenue to crowdsource and process data (i.e. Digivol and Zooniverse)
- may not be usable for some participants.
- assist citizen science projects and enhance an individual's engagement with nature by working as identification guides, tutorials, remote sensors and data collection tools
- may not be useful in remote locations or for people that do not have or know how to use technology.
- Web portals
- Connectivity and transparency of data repository
- It is important to note the location of the data repository and if the data can be extracted so that the collected data are available to other projects.
- Ongoing data hosting or maintenance costs (web portals and apps).
The aim of the project and how frequently data are collected can determine how best to analyse your data. You should consider:
- whether the data collected are quantitative, qualitative or a combination of both
- who will be analysing the data (i.e. expert, participant or both), and if participants will be supporting data analysis additional training may be required to minimise bias and error
- how the data will be used (e.g. understanding the expectations of those you wish to inform) and the acceptance criteria for the data collected.
Identifying individuals in your project team that have data collection, management and analysis expertise will help ensure the scientific rigour of your data and its value for the project and other researchers outside of your project.
Ensure the quality of your data
Citizen science provides the unique opportunity to crowdsource large amounts of data. However, to ensure the collected data are robust, it is important to minimise chances for error and to understand that data quality can differ between participants. Thoroughly test the project's protocols and participant capabilities so that errors in interpretation, measurement and identification can be noted and quantified early.
Consider providing training, developing data collection resources and standards for participants, and creating validation protocols (e.g. participant–coordinator feedback systems when outlier data are noticed). This should help moderate and eliminate the collection of bias data.