Skip to main content
. 2022 Mar 22;10:e13142. doi: 10.7717/peerj.13142

Table 1. Akaike’s Information Criterion (AIC) model selection results for evaluating factors influencing the number of avian collisions detected at two buildings, corrected for differences in the area of glass between the two buildings.

Model structure AIC ∆AIC w df
A treat.grp + season + (1|year) + (1|obs) 335.95 0 1.00 9
B season + (1|year) + (1|obs) 349.63 13.68 0 6
C building + season + (1|year) + (1|obs) 350.25 14.30 0 7
D (1|year) + (1|obs) 352.78 16.83 0 3

Note:

Fixed effects: treat.grp = building by treatment period: Science Complex conventional glass pre-treatment (2013–2015), Science Complex FF post-treatment (2016–2018), Annex ORNILUX® (2013–2015), Annex ORNILUX® (2016–2018); building = Science Complex or Annex; season = Winter, Fall, Spring or Summer. Random effects: 1|year = monitoring year, 1|obs = observer. Offset = window area ∆AIC = change in AIC relative to top model, w = Akaike weight, and df = degrees freedom.