I want to call attention to a graph I published yesterday (source here) showing the relationship between asphalt use in Finland and asthma rates.
There is something very misleading, even wrong, about this graph: the asthma rate is plotted for 20 years longer than asphalt use. This causes several problems.
1. It tricks you into thinking that the asthma/ashphalt correlation has gone on much longer than it really has.
2. It makes the correlation look much tighter. Cover up the area of the graph from 1990 on and the dip in asphalt use between 1980 and 1990 is much more prominent.
Why would you ever present data like this? I can think of no reason to continue graphing one variable and not the other, especially for a period taking up 1/4 of the whole graph. The relationship is a good one without the extra data and this just blows your credibility in the eyes of a careful reader. I also can't help but wonder what asphalt use really was in 1995 and 2000. Maybe it plateaued?