### More School Spending - Summary

A more thorough examination of state school spending data versus

performance shows some minor changes but the overall trend is the same as in my previous blog post.

Method:

(1) State per-pupil spending was calculated as the average of three years (2002-2003, 2003-2004, and 2004-2005) rather than the single year of the previous blog post. This was to capture some of the effects of time; for example, a student's performance is the result of several years of education, not just the one year in which the test is administered.

(2) Performance was taken from the 2005 administration of the NAEP, rather than the 2003.

(3) Reading and math were both examined (the previous blog post was math only).

(4) As before, this is 4th grade data only. I'll look at the 8th grade data another time.

The results (graphs to come later):

Reading, 51 jurisdictions (states plus D.C.): r=.34

Reading, 50 states only : r=.49

Math, 51 jurisdictions: r=.24

Math, 50 states only: r=.40

Range of average reading scores (0-500 possible): 191-231

Cut scores: Basic 208, Proficient 238, Advanced 268

Range of average math scores (0-500 possible):211-247

Cut scores: Basic 214, Proficient 249, Advanced 282

As in the previous data, the scores cluster pretty tightly. So even in Reading, where the correlation between spending and performance is the highest, greater spending will at best lead to modest real world performance increases. Those performance increases could be real and measurable, but not necessarily impressive compared to the amount of money spent to obtain them.

performance shows some minor changes but the overall trend is the same as in my previous blog post.

Method:

(1) State per-pupil spending was calculated as the average of three years (2002-2003, 2003-2004, and 2004-2005) rather than the single year of the previous blog post. This was to capture some of the effects of time; for example, a student's performance is the result of several years of education, not just the one year in which the test is administered.

(2) Performance was taken from the 2005 administration of the NAEP, rather than the 2003.

(3) Reading and math were both examined (the previous blog post was math only).

(4) As before, this is 4th grade data only. I'll look at the 8th grade data another time.

The results (graphs to come later):

Reading, 51 jurisdictions (states plus D.C.): r=.34

Reading, 50 states only : r=.49

Math, 51 jurisdictions: r=.24

Math, 50 states only: r=.40

Range of average reading scores (0-500 possible): 191-231

Cut scores: Basic 208, Proficient 238, Advanced 268

Range of average math scores (0-500 possible):211-247

Cut scores: Basic 214, Proficient 249, Advanced 282

As in the previous data, the scores cluster pretty tightly. So even in Reading, where the correlation between spending and performance is the highest, greater spending will at best lead to modest real world performance increases. Those performance increases could be real and measurable, but not necessarily impressive compared to the amount of money spent to obtain them.