Skip to main content

Table 1 Bayesian growth curve modeling results for boys and girls

From: Bayesian probabilistic forecasting with large-scale educational trend data: a case study using NAEP

  Estimate Post.SD HPD.025 HPD.975 PSRF Prior
Boys growth params.
 Intercept 274.782 1.300 272.116 277.226 1.000 dnorm(260,.1)
 Slope 0.931 0.108 0.726 1.149 1.000 dnorm(0,1e−2)
 Pre(intercept) 88.460 19.415 55.216 127.007 1.000 dwish(iden,3)
 Pre(slope) 0.246 0.067 0.132 0.385 1.000 dwish(iden,3)
Girls growth params.
 Intercept 273.707 1.308 271.113 276.219 1.000 dnorm(260,.1)
 Slope 0.902 0.122 0.669 1.142 1.002 dnorm(0,1e−2)
 Pre(intercept) 85.114 19.065 51.593 122.738 1.000 dwish(iden,3)
 Pre(slope) 0.269 0.084 0.126 0.43 1.000 dwish(iden,3)
  1. pre() refers to the precision of the parameter, where precision = 1/variance. dnorm is the normal \(N(\mu ,\tau ^2)\) distribution, where \(\tau ^2\) is the precision. dwish is the Wishart \((R,\nu )\) distribution for the precision matrix with shape matrix, R and degrees-of-freedom, \(\nu\)