Heavy Lifting - thoughts and web finds by an economist
     I also contribute to Division of Labour Load HL's Front Page
Tuesday, November 28, 2006

NCLB and NYC school attendance

I came across some attendance figures for the NYC school system. The data reflect reported attendance by grade from the 1995-1996 school year thought the 2005-2006 school year. I grabbed the data and threw it into STATA (data file here).

Here's the time trends for grades 7 through 12 over the period in question.



It is amazing how much attendance drops in the ninth grade. Is this when the first cohort of dropouts occurs? Is this when junior figures out how to ditch class and has his first "taste of freedom" to do so?

Anyway, my initial thought was whether there had been a discernible change in average attendance rates in the various grades after the No Child Left Behind Act of 2002. I created a dummy variable that took a value of one after 2002 and zero before. I estimated a fixed effects model with class and year dummy variables:
Fixed-effects (within) regression               Number of obs      =       168
Group variable (i): classid Number of groups = 14

R-sq: within = 0.2026 Obs per group: min = 12
between = 0.0000 avg = 12.0
overall = 0.0284 max = 12

F(1,153) = 38.88
corr(u_i, Xb) = -0.0000 Prob > F = 0.0000

------------------------------------------------------------------------------
attendance | Coef. Std. Err. t P>|t| [95% Conf. Interval]
-------------+----------------------------------------------------------------
nclb | 1.482143 .237701 6.24 0.000 1.012543 1.951743
_cons | 88.67857 .1372367 646.17 0.000 88.40745 88.9497
-------------+----------------------------------------------------------------
sigma_u | 3.9913721
sigma_e | 1.4523771
rho | .88307391 (fraction of variance due to u_i)
------------------------------------------------------------------------------
F test that all u_i=0: F(13, 153) = 90.63 Prob > F = 0.0000
The model doesn't explain much (but it wasn't intended to). The test is basically for a difference in average attendance before and after the NCLB. It looks like the average attendance rate increased about 1.4 percent after NCLB. This, of course, does not prove causation, far from it.

However, it suggests that further investigation might be warranted. Regardless of whether NCLB has the intended impact on test scores, teacher and administrator accountability, etc., perhaps it would be enough if students and parents expect the policy to have an impact. This might encourage increased attendance (and less dropouts?), at least in the short run.

Just a passing thought.

Comments: Post a Comment



Purchase




Le Chai - galerie du vin



Support



Popularity

Posts that contain Craig Depken per day for the last 90 days.



Contacts

Heavy Lifting's Main Page
Email Me
Atom Feed

WWW
Heavy Lifting




Great Links

Money I Found Today

Heavy Lifting - Firehose style (56k warning)



Recent Posts

- NCLB and NYC school attendance




Archives


home


Visitors
Site Meter Blogroll Me!



Credits

Modified maystar design
powered by blogger