There seem to be some interesting things happening over on the X-Y scattergraph with a good correlation appearing in the school size graph. Whilst there seems to be little correlation occurring between any of the other variables and illness rates, there does seem to be a positive correlation with school size – it’s possible that larger schools have higher rates of absence-due-to-illness.
There is a fairly good correlation between school size and illness absence during the second half of November and early December while absence levels are still broadly fluctuating around the background (and as such you would expect the greatest error in the results). The really interesting thing that seems to happening at the moment is that this correlation disappears during the peak of the illness absence just before Christmas when we might have otherwise expected the strongest correlation.
This could be down to chance or any number of confounding factors in the last week of term. The only way to know that what we’re seeing is genuine is to get more data points:
I for one would be really interested in hearing people ideas about what might have caused this collapse in correlation in the peak weak.