Data-derived metrics describing the behaviour of field-based citizen scientists provide insights for project design and modelling bias

Abstract

Around the world volunteers and non-professionals collect data as part of environmental citizen science projects, collecting wildlife observations, measures of water quality and much more. However, where projects allow flexibility in how, where, and when data are collected there will be variation in the behaviour of participants which results in biases in the datasets collected. We develop a method to quantify this behavioural variation, describing the key drivers and providing a tool to account for biases in models that use these data. We used a suite of metrics to describe the temporal and spatial behaviour of participants, as well as variation in the data they collected. These were applied to 5,268 users of the iRecord Butterflies mobile phone app, a multi-species environmental citizen science project. in contrast to previous studies, after removing transient participants (those active on few days and who contribute few records), we do not find evidence of clustering of participants; instead, participants fall along four continuous axes that describe variation in participants' behaviour: recording intensity, spatial extent, recording potential and rarity recording. Our results support a move away from labelling participants as belonging to one behavioural group or another in favour of placing them along axes of participant behaviour that better represent the continuous variation between individuals. Understanding participant behaviour could support better use of the data, by accounting for biases in the data collection process. Human activities are causing changes in biodiversity at both global and local scales 1. Climate change, land use change, and globalisation have led to wide scale loss of biodiversity 2 and the translocation of species around the globe 3 , with negative impacts on ecosystem functioning 4,5. Citizen science offers a mechanism to monitor the status and change of biodiversity, identify drivers of change, and report on progress in reversing declines 6. Citizen science engages volunteers and non-professionals in scientific research 7 and has grown and diversified rapidly in recent years, due in part to technological advances 8,9. It provides benefits both in the engagement of people and in the cost-effective collection of environmental data 10 and so offers a mechanism for large scale and long term data collection that would not be practical otherwise. Citizen science data are increasingly used to understand large scale processes in ecological and conservation biology such as the impacts of climate change 11,12 , pesticides 13 , invasive species 14,15 , and poaching 16. The diversity of citizen science projects is exemplified by projects that focus on the recording of wildlife observations. These projects range from highly structured recording schemes that require taxonomic expertise to participate, to simple opportunistic projects that encourage mass participation 17-19. Allowing flexibility in the data collection process allows participants to choose how they engage with a project. This has...

Citizen science for monitoring seasonal-scale beach erosion and behaviour with aerial drones

“…Bias in citizen science projects can be introduced by allowing individuals flexibility in how, when and where to collect data 47 . We reduced the risk of bias prior UAV operations by (1) assigning groups to fixed locations, (2) targeting low-tide for survey, (3) providing standardised protocols and training on simple to use highly automated equipment and (4) assisting citizen scientists at each sites for the first three flights.…”

The role of passive surveillance and citizen science in plant health

“…Research can contribute greatly to inform the standardisation of methods and to improve data analysis. Where observers are numerous and providing repeat observations methods from species distribution modelling can be applied to understand bias and improve model accuracy (August et al 2020 ; Isaac et al 2014 ; Renner et al 2015 ). For plant health issues, it remains particularly important to understand the components that affect detection (namely the probability to detect, probability to identify and probability to report the issue), so that inferences can be made quickly when new threats are discovered.…”

“…These reports occur at times and places where it is convenient for the public to participate and, as such, are unstructured and potentially biased in distribution (Isaac and Pocock 2015 ; Baker et al 2018 ; Johnston et al 2020 ). This systematic error contravenes a fundamental assumption of most statistical approaches (that data is a representative random sample of the wider population) (Dobson et al 2020 ; August et al 2020 ). For analysts and policy makers there are clear limitations to how these data can be used, but there is no denying the practical usefulness of additional detections to plant health officials (Ryan et al 2018 ).…”

“…For analysts and policy makers there are clear limitations to how these data can be used, but there is no denying the practical usefulness of additional detections to plant health officials (Ryan et al 2018 ). It remains an important challenge for researchers to develop methods for use with passive surveillance data, so that any bias can be understood and quantified (August et al 2020 ; Boakes et al 2016 ).…”

A novel approach to the classification of terrestrial drainage networks based on deep learning and preliminary results on solar system bodies

“…The complexity of the human intelligence emulation of natural phenomena by data-driven systems highlights the need to acquire knowledge, extracting conceptual schemes and complex data correlations directly from the primary source, which is the observed data of the studied phenomenon [39][40][41] .…”