I found this interesting:<p>>The article describes the conversion of Mountlake Terrace High School<p>> in Seattle from a large suburban school with an enrollment of 1,800<p>> students into five smaller schools.<p>> The conversion was greased with a Gates Foundation grant of almost a<p>> million dollars. Although <i>class sizes remained the same</i>, each of the<p>> five schools had <i>fewer teachers</i>.<p>This is really funny as there is no really sensibly put hypothesis that why the smaller schools should fare better.<p>So they looked at data and thought that they see that smaller schools do better but forgot to ask "why?".<p>Now having hypothesis that smaller classes make results better as teachers would have more time to deal with each child makes much more sense.<p>And this could be then tested.
Case in point: Of the ten most dangerous cities in California, seven are small towns such as Nevada City (pop. ~3,000). The other three are Emeryville, Stockton, and Oakland.<p><a href="http://www.telegraphtoday.com/10-most-dangerous-cities-in-california/" rel="nofollow">http://www.telegraphtoday.com/10-most-dangerous-cities-in-ca...</a><p>De Moivre must be spinning in his grave.