The main results that emerged from the present thesis were: (1) cognitive load affect citizen generated data quality of pollinator monitoring: (2) bumblebees and butterflies are in general easier to identify in comparison to syrphid flies and solitary bees both digitally and in field: (3) students generate data with higher quality than children.
Students found it harder to correctly detect a butterfly when the number of pre-existing insects increased in a square. This is consistent with the cognitive load theory and thereby not surprising. Interestingly, students found it easier to identify bumblebees when the number of pre-existing insects in a square increased. This contradicts the cognitive load theory, however, humans have preference for objects in canonical size (their true size). The bumblebees in the video recordings were true to their live size and were larger in comparison to the pre-existing insects in the square, which might have led to bias for bumblebees.
Both students and children struggled to correctly detect solitary bees. When the number of solitary bees increased in a segment, students found it approximately three times harder to correctly detect and children found it approximately six times harder. Even if it was hard for both students and children, my results show it is a bit harder for children. This might be due to children having a low attention-holding span and focusing on other things.
Interestingly, children found it easy to correctly detect a syrphid fly when its proportion increased in a segment while students found it hard to correctly detect a syrphid fly. Syrphid flies are Batesian insects (they mimic predators) and it is common they mimic wasps (Moore & Hassall, 2016) and according to Golding and collaborators (2005), syrphid flies mimicry fools’ humans. The reason why children found it easy to correctly detect syrphid flies might be due to children’s ability to separate multiple stimuli and due to that, the children could separate syrphid flies’ morphological differences
Another colorful insect order is butterflies and it is no surprise that the children thought it became easier to correctly detect a butterfly when its proportion increased in a segment. What is interesting is that students found it harder to correctly detect a butterfly when its proportion increased in a segment. During a digital F.I.T-count it gets hard to know if you have recorded an individual before. The segments where butterflies were supposed to be detected were crowded with other butterflies flying and since the students needed to keep track over if they have reported an individual or not, they suffered inattentional blindness (when preoccupied with something else you miss what is in front of you).
Both children and students found it easier to correctly detect a bumblebee when its proportion increased in a segment. It was approximately two times easier for both groups to correctly detect bumblebees. This is not surprising since citizen science is already established in bumblebee monitoring and provides scientists with useful data
In conclusion, citizen science has potential for research at spatial scales, but data quality should be considered carefully. I suggest that citizen science can be an option for pollinator monitoring at the level of order, however more training and engagement is necessary for increased data quality for syrphid flies and solitary bees.