A Markov decision model for at-risk schools in the United States under 'No Child Left Behind'

摘要

The new federal law of No Child Left Behind (NCLB) seeks to close student achievement gaps through flexible educational approaches and supporting parent's rights especially for children in lower socio-economic and culturally diverse populations. However, the measure of success under NCLB has been debated because it requires the schools, especially "at risk" schools to reverse the academic performance bands. However, the records of the past years show that school improvement has been linear with a bidirectional and variable rate of change. In order to satisfy the ultimate goals of NCLB, which aims at all students to perform at or above proficient, a non-linear Markov model has been proposed for student movement through achievement levels from performance bands of 1. FBB, far below basic, 2. BB, below basic, 3. B, basic, 4. P, proficient, and 5. A, advanced. The results of the model in terms of prediction for 12 years have been presented using three simulation scenarios. (6 refs.)