Thursday 16th February
Speaker: Jonathan Silvertown | website
Theme: Spectrum of Citizen Science
Citizens pay for all our science, if we allow them to engage or not, which is a reason for us as scientists not only to respond to their wishes but to engage them in science as well. It is useful to share experience on the projects we have been involve in.
Think about what people know and how we use this knowledge. One can distinguish between three types of crowd knowledge:
1) Wisdom of the crowd; everyone’s opinion has the same weight, decisions made by vote. Network typology: equal weights.
2) Wisdom in the crowd; opinions are weighted by expertise. Network typology: experts
3) Wisdom from the crowd; synergistic interactions –> whole > sum of parts. Network typology: synergy.
EvolutionMegaLab is an example of the first type of crowd knowledge, where everyone’s opinion has the same weight. A project where people collect information on snails of one population. The project set out to test the hypothesis that climate change has influenced the morph ratio’s of the snails. The colour of the snails shell shows the temperature, so using a longitudinal study across Europe, one can test this hypothesis with the public quite easily. A major advertising campaign kick started the project and the project website is available in 14 languages. The data is being translated into pie charts layered on maps. Asked people to go back to certain area’s, providing detailed instructions in local languages. The recording sheets are filled in on the website and submitted, the software gives immediate feedback by making a comparison to other data within 5 km radius of yours.
How do we know that users are correctly identifying the different morphs? Train users by an online quiz to train people to classify the morphs, use the quiz results to weigh their data by score. Problem: the majority of people didn’t bother to take the quiz. The results of right/wrong were between 33-95% correct, depending on the question. Quiz results indicated that juvenile C.nem might have been mistaken for adult C.hort (only 33% correct) so the data on C.hort has been omitted.
How much did it cost and is it cheaper to get the public to do it instead of doing it yourself? No, not cheaper, but not only about just collecting data, have public engagement in climate change and exposure to scientific methods.
Geographic spread of data from public is different from that collected by scientists before because the public tends to collect data close to the areas they live (cities) whereas biologists enjoy taking fieldtrips to the country side to collect data.
3 Lessons from EvolutionMegaLab: 1) Difficult to evaluate the skill of every participant. 2) But indirect (quiz) evidence can be used in verification. 3) Creating a self-help network among users might improve data quality.
iSpot is not a classic citizen science project, but more a tool for citizens to create the new generation of naturalists. Need to put names to biological observations because without it cannot say if the species are increasing or decreasing etcetera. iSpot fits the expert based typology of network in a sense; One person is a beetle expert, a fungus expert, a plant expert etcetera, No one is an expert on all so putting these together means that in total you actually have a synergistic interaction. Use reputation of people, built in methods for people to earn points and move up the learning curve. iSpot has been highly succesful. Within an hour over 50% of the observation have been ID-ed by the online community. When the number of observations per month goes up, the percentage of observations with ID stays the same.
The next big challenge is: what do we do when iSpot goes global? How do you globalize expertise? The answer is already there in the network.
Lessons from iSpot: 1) Reputation, its recognition and the scope to increase it through learning drives iSpot 2) Multiple roles as expert/beginner work
Citizen science is as much about the sociology as it is about the data.
Extreme Citizen Science blog 