Suggestions for Common Data and Model Structures for Island Closures Analyses

• Main Author: Doug Butterworth (10005953)
• Format: Report
• Language: unknown
• Published: 2021
• Subjects: Biological Mathematics; Applied Statistics; Biostatistics; penguin island closure experiment; data and model structures; island closure analyses; Penguin Island
• Online Access: https://doi.org/10.25375/uct.14186165.v1

Following the statement of the recent International Review Panel meeting concerning the desirability of the use of common data sets and model structures for comparative analyses of data from the island closure experiment, suggestions are offered in both regards. In terms of agreed data protocols for the experiment, the datasets held by DEFF need to be updated up to and including 2018, and to include information on the month co-variate, where this is possible. For chick survival, discussion is needed on how to proceed, given that different methods of analysis have been suggested to transform the raw data into the form needed for input to the model used to estimate the island closure effect. For model structure, the default agreed locally and by the Panel in 2016 is suggested. Associated aspects related to the chick survival and especially the chick condition response variable are raised as requiring further discussion.