Effects of Ecological Complexity on Student Identification Accuracy in a School-Based Citizen Science Program

Area of Science:

• Ecology
• Citizen Science
• Biodiversity Monitoring

Background:

• School-based citizen science is a growing tool for biodiversity monitoring.
• The reliability of ecological data generated by students is often uncertain.

Purpose of the Study:

• To evaluate the reliability of student-generated ecological data from a nationwide barn owl pellet project in Israel.
• To assess student identification accuracy and factors influencing it, such as sample complexity and prey taxa.

Main Methods:

• A nationwide barn owl (Tyto alba) pellet project involving 1612 students and 107 teachers over two academic years.
• Analysis of 3333 prey records generated from returned owl pellet kits.
• Assessment of identification accuracy, including perfect identification rates, item-level accuracy, and the impact of pellet complexity.

Main Results:

• Teacher participation was moderate to high, with schools contributing significant sampling effort across broad geographic areas.
• Student identification accuracy was moderate, with 50% of pellet analyses perfectly identified and an overall item-level accuracy of 62%.
• Identification reliability decreased with increasing pellet complexity, with each additional prey item reducing the odds of perfect identification by approximately 68%.

Conclusions:

• School-based citizen science can contribute to biodiversity monitoring efforts.
• Reliable ecological inference from student-generated data requires validation frameworks, particularly for complex samples.
• Student engagement remained high despite initial perceptions, with enjoyment being a key factor for continued participation.