Stat of the Week Aug. 10, 2006
Statewide Performance-based Policies Towards Schools
Federal law dictates the use of sanctions for Title I schools not making adequate yearly progress on state tests. Some states have adopted their own policies on sanctioning or rewarding schools based on performance. These policies go beyond federal requirements in that they apply to all schools in the state and may combine both sanctions and rewards.
In this week’s Stat of the Week, the EPE Research Center looked at how many states had such performance-based policies in place in 2003 and at where those states were located. Next, gain scores on the fourth grade NAEP math test were calculated between the years 2003 and 2005 for groups of states based on the type of policy in place or the absence of such a policy.
In the school year 2002-03, slightly over half (53%) of all states had in place some kind of a performance-based policy towards schools. Looking at those states alone, 15% issued rewards only, 37% issued sanctions only, and 48% issued both sanctions and rewards.
Generally speaking, states in the southern half of the country, especially in the Deep South, were much more likely to have had performance-based policies in place than those in the northern states, with the exception of a cluster of states in the Northeast. The accompanying map depicts the locations of states by type of policy.
Looking at the fourth grade NAEP math test scores reveals that states with both reward and sanction policies in place had the lowest test scores in both 2003 and 2005 while states with no policy in place had the highest test scores both years (although the differences are small.)
Math achievement gain scores over the two year period revealed that states with a policy that included rewards (either in addition to a policy for sanctions, or by itself), showed slightly greater gains on the fourth grade NAEP math test than did states that offered sanctions alone (an average of 3.4 for the two reward groups versus 2.8 for the sanctions-only group). When the two groups of states offering rewards were compared to states that had no policy at all, the gain score difference was smaller although it still favored the reward policy states (3.4 versus 3.2).