You would think that

if you thought time of day wasn't a factor.

Since these are random shootings, you would expect women and girls to be victims at roughly their proportion of the population, which is slightly over 50%--or maybe higher, if it turns out that the shootings are concentrated in urban areas with low sex ratios. The question, then, is not why so many victims of this type of shooting are female but why a disproportion of them (55.2%) are male.

From James Taranto's "Best of the Web" column at the Wall Street Journal today.

Ockham's Razor does not ask us to ignore context. Taranto gets the numbers thinking right as far as he goes. But he fails to ask if time is a factor in the random shooting of innocents. 

Let's see the Sex*Time of Day (by age) graph. I suspect that for kids younger than 18 or so, there's a statistically significant sex difference. 

Compare it to the Violent Crime*Time of Day graph. 

Any normal parent is more protective of daughters than of sons. Or should be. Boys get to be outside later than girls.  

Taranto assumes that the sample on the street at the worst time to be on the street is a representative sample of the  population. 

It is not. 

There are more boys than girls.  

UPDATED: At least that's my hypothesis.