Recently there was some research done on Alzheimer’s disease using a game that collated data from a vast number of people. It was revealed that if the research had been done in a clinical, controlled trial it would have taken over 2000 years to collect the same amount of data. This is one of the best and most informative study’s ever done using citizen science, where “regular people” (non-scientists) collect the data and are actively involved.
Unsurprisingly this led to a lot of hype and lots of people calling for more citizen science. All very admirable we can agree but it is not without its problems.
Citizen science is not easy. The Alzheimer study was so good because it had been set it up with citizen science in mind and designed specifically for this purpose. This is a very important point and it led to vast numbers of people being able to do something relatively simple to help. The medical world seems to have a better grasp on this than the world of conservation where citizen science is often dismissed for being poor quality and some key journals are allegedly refusing to publish papers based on citizen science derived data. Some think this is academic snobbery whereas others argue that often the projects are poorly run leading to errors which then cannot be published.
I follow a lot of marine science quite closely and regularly see citizen science crop up in one guise or another. Likely this is due to a lack of funding in this area which means researchers cannot afford full scale studies and must adapt. This is fine, but not all projects are being carried out in a useful way because the experimenters are not designing the project in the knowledge that it will be “normal” people who are doing the data collection. In some cases researchers are only offering advice to groups and then the groups do everything themselves, including publishing the results. The results are not published academically but are made publicly available and are then leading to people having incorrect ideas on the health of marine systems, particularly when the results make local or national news before anyone checks the facts. The purpose of academic publications is to have everything peer reviewed before publication to ensure that the details are correct, to skip this step really calls into question how scientific the project is and how much trust can be placed in its findings.
Marine and coastal science is often an exceedingly difficult area to get data from, seabird surveys for example are 70% accurate at best, sometimes as low as 30% and this leads to a huge issues with data analysis. Many people on this website are students and recent graduates who are all too familiar with statistical analysis and are aware of the degree of confidence required in your results prior to publication.
If you are unfamiliar then try this example. You count 5000 animals in 2015, alter conditions and count 5500 animals in 2016. The population has gone up by 500 and proves whatever you have changed has worked, right?
As you may have guessed that is not necessarily the case. To make that claim you would need to prove that the change wasn’t due to different people counting, different weather, time of year, and a vast, vast, number of other factors. In the case of seabirds if the very best studies are only counting 7 out of every 10 birds then there is clearly huge scope for error, error that needs to be statistically controlled for. But that is only to prove the population change, to prove that what you did caused the population change is a huge leap requiring compelling evidence. This level of analysis needs to be done in all research and yet it is missing in a great deal of citizen science. By no means is that the fault of the citizens, rather the motivation of the researcher(s) or the design of the project is the issue.
The purpose of this is not to shame the groups involved. It is unlikely that anyone is deliberately setting out to mislead, although there have been rare occasions when citizen science projects were created to “prove” certain things in order to influence decisions (a claim that can be levelled at academic research as well) . Some of the time it seems the projects are either without academic support at all or are not a priority for researchers which undermines the amount of work being put it.
Citizen science is not the problem, the problem is that it needs to be recognised as real research and treated as such. Researchers need to commit to their groups, provide any necessary training and design the experiment in such a way as to minimise data collection errors. This is not necessarily more difficult or easier than a fully academic project but it does require a different approach. If citizen science is going to make the same sort of positive contribution to conservation that it can make in other fields then it must be an all or nothing situation where it is done properly or not at all. Given that people are clearly willing to help then designing a project that takes advantage of this is surely the way forward.
There is a good summary of the issues, particularly in the comments section, in the Nature link below.
As always, let me know what you think in the comments below.
1,152 total views, 1 views today