A heat map representation of two data sets used in the artificial mixing experiment.

<p>In each of the heat maps the data collection area was split into 45,576 equally sized rectangular cells. The number of observations that occurred in each of these cells was then enumerated. In both plots this enumeration was done using the data from the entire flock. Cells given in blue correspond to areas in which at least one observation was noted. Yellow and red cells correspond to areas where at least 0.01% and 0.05% of the observations occurred. This roughly corresponds, respectively, to 4.5 and 22.5 times the number of observations that would have been seen had the observations been uniformly spread over the rectangular cells. It can be observed that in both of the individual data sets the flock spends a disproportionately large amount of time in the top right section of the field. Properties such as this, which occur across the individual data sets, may go some way to accounting for the fact that the false-positive rate is slightly higher than expected in the artificial mixing experiment.