Question 2.4

Is it possible that plots have been sampled in a non-random fashion at the start of the national monitoring schemes – because observers tended to count at sites with a high species richness and abundance of birds – biasing the results for the first years?


A non-random selection of plots has indeed occurred in some cases, but in these cases the plots once chosen were consistently monitored during following years. When new, randomly chosen plots were added to the initial plots, we imputed missing values for these newly added plots in earlier years, thereby preventing any bias.

See more in Methods, chapter 1, Box Selection of sample plots and Box Missing values I.