Comments on: Study: The one change districts should make to save $10 billion http://www.eschoolnews.com/2012/09/06/study-the-one-change-districts-should-make-to-save-10-billion/ Just another eSchool Media site Thu, 10 Jan 2013 16:37:10 +0000 hourly 1 http://wordpress.org/?v=3.4.2 By: michaelmflood http://www.eschoolnews.com/2012/09/06/study-the-one-change-districts-should-make-to-save-10-billion/comment-page-1/#comment-165322 michaelmflood Thu, 06 Sep 2012 17:21:31 +0000 http://www.eschoolnews.com/?p=108359#comment-165322 This study is full of flawed methodologies. There was intense pre-selection criteria eliminating 90% of the districts in the US from the study. They spend most of the study looking at 10 "pairs" (20 districts) specifically chosen to illustrate their point. They even state in the study: "(We do not imply that these relationships are causal. And we’re mindful that the district pairs were chosen to illustrate the inverse relationship between special education inputs (spending) and outcomes (achievement)—so it’s not surprising that they did, in fact, illustrate that relationship.)" They took a small response, narrowed it farther, then picked 20 to show a point they decided in advance to make. That isn't science, but I suppose we could save money by not teaching science or statistics either... In their general data analysis, they compare SPED spending & staffing levels to TOTAL ENROLLMENT, not to the # of SPED students. They assume any two districts with the same total enrollment will have the same needs in their population. That is crazy. A district with a world-leading center on the study of autism will have a higher SPED rate than a district without it. This study is written with the "outcome" as a "goal", not in the spirit of discovery or scientific / statistical analysis. It is loaded with sampling flaws and mthodology that is highly suspect. This study is full of flawed methodologies. There was intense pre-selection criteria eliminating 90% of the districts in the US from the study. They spend most of the study looking at 10 “pairs” (20 districts) specifically chosen to illustrate their point. They even state in the study:

“(We do not imply that these relationships are causal. And we’re mindful that the district pairs were chosen to illustrate the inverse relationship between special education inputs (spending) and outcomes (achievement)—so it’s not surprising that they did, in fact, illustrate that relationship.)”

They took a small response, narrowed it farther, then picked 20 to show a point they decided in advance to make. That isn’t science, but I suppose we could save money by not teaching science or statistics either…

In their general data analysis, they compare SPED spending & staffing levels to TOTAL ENROLLMENT, not to the # of SPED students. They assume any two districts with the same total enrollment will have the same needs in their population. That is crazy. A district with a world-leading center on the study of autism will have a higher SPED rate than a district without it.

This study is written with the “outcome” as a “goal”, not in the spirit of discovery or scientific / statistical analysis. It is loaded with sampling flaws and mthodology that is highly suspect.

]]>