This comment reanalyses the data presented in Rydell (Rydell 2017 R. Soc. open. sci.4, 161077. (doi:10.1098/rsos.161077)) which were analysed using only very basic statistics like Fisher's exact test and McNemar's test. We demonstrate how the use of more advanced statistical methods can make better use of the available data, quantify the observed effects and strengthen the conclusions in Rydell (Rydell 2017 R. Soc. open. sci.4, 161077. (doi:10.1098/rsos.161077)). We have no intention to discredit the original authors. Their analyses were basic but correct.
Keywords: biodiversity; cultural heritage; historic buildings; light pollution; mixed models; ordinal regression.