Schools and Inequality: A Multilevel Analysis of Coleman's Equality of Educational Opportunity Data - 2 views

• 
  #1 Bonnie Sutton on 27 Jan 12

   

  by Geoffrey Borman & Maritza Dowling
  Four decades after the pathbreaking Coleman report, researchers are still working to address its primary message: that school social composition and resources are not important for understanding and addressing educational inequality. Using the original Equality of Educational Opportunity data, this study applied a two-level hierarchical linear model to partition the variation in ninth-grade students' verbal achievement into its within- and between-school components and to measure the associations among school-level social composition, resources, teacher characteristics, and peer characteristics and achievement. We estimated that 40% of the achievement variance was between schools, whereas Coleman and colleagues had originally estimated that only 8.5%-18% lay between schools. Explanatory analyses suggested that the racial/ethnic and social class composition of a student's school was over 1 3/4 times more important than a student's individual race/ethnicity or social class for understanding educational outcomes. Further, within-school Black-White achievement gaps and social class differences were explained in part by curricular differentiation and teachers' preferences toward middle-class students. These findings are contrasted with those from a set of traditional ordinary least squares regression models and the past conclusions drawn from the Coleman report. This is a long research paper.
  http://www.tcrecord.org/Content.asp?ContentID=15664

   

  <div class="cArrow"> </div><div class="cContentInner">by Geoffrey Borman & Maritza Dowling <br /> Four decades after the pathbreaking Coleman report, researchers are still working to address its primary message: that school social composition and resources are not important for understanding and addressing educational inequality. Using the original Equality of Educational Opportunity data, this study applied a two-level hierarchical linear model to partition the variation in ninth-grade students' verbal achievement into its within- and between-school components and to measure the associations among school-level social composition, resources, teacher characteristics, and peer characteristics and achievement. We estimated that 40% of the achievement variance was between schools, whereas Coleman and colleagues had originally estimated that only 8.5%-18% lay between schools. Explanatory analyses suggested that the racial/ethnic and social class composition of a student's school was over 1 3/4 times more important than a student's individual race/ethnicity or social class for understanding educational outcomes. Further, within-school Black-White achievement gaps and social class differences were explained in part by curricular differentiation and teachers' preferences toward middle-class students. These findings are contrasted with those from a set of traditional ordinary least squares regression models and the past conclusions drawn from the Coleman report. This is a long research paper. <br /> <a href="http://www.tcrecord.org/Content.asp?ContentID=15664" rel="nofollow" target="_blank">http://www.tcrecord.org/Content.asp?ContentID=15664</a></div>

  ...

  Cancel

To Top