Inter-Observer Reliability for Large Exotic Felids
Abstract
The objective of this study was to determine and refine the inter-observer reliability scores before a large exotic felid enrichment study was conducted in a zoological setting. Three felids housed at Omaha’s Henry Doorly Zoo & Aquarium were used. An ethogram was adapted from previously published felid work. Three behavioral measures were compared; active, inactive and other. A total of 6 female observers were used. One trainer (Observer 1 = control) who had one year of behavioral research experience with large exotic felids was responsible for observer training prior to study commencement. To test for inter-observer reliability, three 10 minute videos were viewed (sessions 1 through 3). All observers independently reviewed these videos. Each observer scored the felids using a one minute scan sampling technique. For each session there were 10 data points collected per observer or 30 data points per observer over the three training sessions. Data will be presented descriptively. There were two data points from observer 5 that differed from Observer 1 (control). There was one data point from Observer 6 that differed from Observer 1 (control). Therefore, 177 / 180 data points were in agreement between Observer 1 (control) and observers 2 through 6. This resulted in an overall inter-observer agreement of > 98% before the experiment began. In conclusion, when using multiple observers for a behavioral study it is critical to have highly skilled individuals so that the data collected is accurate so that any treatment effects can be determined.
How to Cite:
Iske, C., Morris, C. L. & Johnson, A. K., (2016) “Inter-Observer Reliability for Large Exotic Felids”, Iowa State University Animal Industry Report 13(1). doi: https://doi.org/10.31274/ans_air-180814-256
Downloads:
Download pdf