dateduc: Student test scores and associated teachers for a single school in a large urban school district
Description
Response captures vertically linked mathematics and reading scores on a norm-referenced
standardized test administered during the spring of the years 1998 to 2002 obtained from
a large urban school district. Included student are in grade 1 during the 1997 - 1998 school
year and followed successively until grade 5 for the 2001-2002 school year. These data focus
on a single school of 227 students and 34 teachers. Teacher links to students did not vary
by subject. The data are configured to support model runs using engine functions dpgrowmm and ddpgrow.
Format
A list object for 562 total observations on 227 subjectsDetails
- y. Numeric vector of
N = 562 student-year test scores.
- subject. Numeric vector,
1, ..., n = 227, student identifiers
- trt. A numeric vector of
N 0's to indicate there are not separate treatment arms.
- time. An N x 1 numeric vector of associated meaurement times in
1-5.
- subj.aff. Same as
subject as all students receive the "school" treatement.
- W.subj.aff. An
n = 227 x S = 34 matrix object that links the n = 227 students to the S = 34 teachers. The rows sum to 1.
This object is used in engine function dpgrowmm.
- dosemat. An
n = 227 x S = 34 numeric matrix where the first column is 1's for the intercept and the first teacher is left out for identifiability.
This matrix is the same as W.subj.aff except that the first column is replaced with an intercept. This object is use for engine function ddpgrow.
- tchr.aff. A numeric vector with values
1,...,T=34 encoding the id's for the participating teachers.
- labt. A character input providing the label
"school" for the treatment delivered to students
References
J.R. Lockwood, Daniel F. McCaffrey, Louis T. Mariano and Claude Setodji (2007) Bayesian Methods for Scalable Multivariate
Value-Added Assessment, Journal of Educational and Behavioral Statistics, 32(2), 125 - 150.S. M. Paddock and T. D. Savitsky (2012) Bayesian Hierarchical Semiparametric Modeling of Longitudinal Post-treatment Outcomes from Open-enrollment Therapy Groups, invited re-submission to:
JRSS Series A (Statistics in Society).