Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

st: new package -vam- available in SSC

From   Michael Stepner <>
Subject   st: new package -vam- available in SSC
Date   Tue, 8 Oct 2013 00:29:56 -0400

-vam- is a Stata routine which computes drift-adjusted teacher
value-added measures.  It is now available in the SSC, with thanks to
Kit Baum.

vam uses student-level scores to compute teacher value-added (VA)
measures. For each teacher, vam estimates the best linear prediction
of value-added in each observed year, based on the scores of students
taught by that teacher in other years (prior and future).

vam does not impose the restriction that each teacher's VA is fixed
over time, which is an important departure from standard VA models.
The estimation procedure accounts for drift in teacher VA by
permitting the coefficients on score data to vary non-parametrically
according to the distance between each observed score and the forecast
year.  In practice, drift has been observed to be is econometrically
important in long panels, where test scores from recent years are a
better predictor of current teacher quality than more distant lags.
See Section 1 of the paper cited below for further discussion.

This program implements the estimation procedure developed by Chetty,
Friedman and Rockoff (2013), described in Appendix A of their paper:

Chetty, Raj, John N. Friedman, and Jonah E. Rockoff. 2013. "Measuring
the Impacts of Teacher I: Evaluating Bias in Teacher Value-Added
Estimates." NBER Working Paper 19423.

You can install the program via -ssc install vam-.

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index