Of all the states, Maryland has the smallest "N-size" by far.
N-SIZE: STATES
----------------------------------------------------------
05 : MD
10 : LA SD UT
11 : NH
16 : CO
20 : ME MN
25 : AK AR NM NV
30 : FL IA IN KS MI MO MT NE NJ NY OH OK PR WA WY
34 : ID
40 : AL AZ CT DE GA HI MA MS NC PA SC VT WI
42 : AZ OR
45 : IL RI TN
50 : CA TX VA WV
Missing: KY and ND
You can also find N-sizes for many states on this "accountability illusion" data map.
If I remember my statistics correctly, statisticians recommend--in most situations-- a sample size no smaller than 30 or 40 to have reasonable confidence that the sample is representative of a population.
Almost all of the states were in this ballpark.
My guess is that the folks from the Maryland Department of Education who chose the tiny N-size of five would flunk Statistics 101 and perhaps get forty lashes from their statistics teacher. Nancy Grasmick, are you listening?
To find the reasoning behind your state's N-size, go here, open the PDF for your state's "Consolidated State Application Accountability Workbook" and scroll about two-thirds of the way through the document to "critical element" #5.5.
2010-03-21
Minimum "N sizes" chosen by states for NCLB AYP reporting
Labels:
accountability illusion,
adequate yearly progress,
AYP,
N size,
NCLB,
schools
Subscribe to:
Post Comments (Atom)
If you took the time to reearch what the "n" represents in this instance you might actually begin to understand that Maryland has set a higher standard than all other states. the subgroup size means that we can't discard struggling children. When you actually get a clue, your comments may be more valid.
ReplyDeleteI'm happy to stand corrected if I'm wrong. I *did* take a good bit of time to research it.
ReplyDeleteCan you explain from a statistician's point of view why such a small "n-size" makes sense?
In this case the larger the "n" - the larger the group of students that school systems can "discard" - in other words, if you want to make AYP easily you assign a larger "n" size. In the case of Maryland schools - the smaller "n" holds schools more accountable for educating all children - if the subgroup is 5 and they don't make it - then the AMO is not met for that group.
ReplyDelete