This map shows the difference of the 2016 rate of non-primary residence dwellings from our modelled prediction. We model the rate of non-primary residence dwellings as being given by the respective 2011 rate in each area plus a rate of 10.1% for each net new unit. (10.1% is the Canadian average rate of non-primary residence buildings for each net new unit, assuming the rate is unchanged for all other dwelling units.)

While this map is easily interpretable, it suffers from sampling bias. Areas with lower number of dwellings have a higher statistical chance of being an outlier. When zooming into the dissemination block level the map is dominated by statistical noise. To extract useful information we need to click into each region and examine the number of dwelling units to understand the reliability and significance of the change.

Our Bayesian surprise map avoids this by colouring by probability that the model is consistent with the observed values, rather than colouring directly by the difference of observations and model.