Answer: Fewer impoverished students.
As a result of my analysis of academic outcomes between charter and traditional public schools (found here), I was asked to perform an analysis into to why the reading scores in charter schools were significantly higher. The results of that analysis can be found below.
Working under the hypothesis that the number of that student socioeconomic status (negatively associated) and charter school status (positively associated) are predictors of students’ reading test scores, I conducted a multiple linear regression analysis.
A correlational analysis confirmed significant associations between these factors consistent with my hypothesis. The percentage of students from low-income households is negatively associated with reading scores, r(2,593) = -.72, p < .001. A weak, but positive, positive association between charter school status and reading scores with also exists, r(2,593) = .13, p < .001.
For both of the predictor values analyzed (SES status and charter school status), their combined effect on the reading scores explains approximately 52% of the total variance, R² = .5246, F(2, 1952) = 1076.92, p < .001. Regression analysis also found that student SES is negatively associated with reading scores, b = -.37, t (1,952) = -45.65, p < .001, with every one-point increase in a school’s’ percentage of students from low-income households the school’s’ average reading score will decrease by .37 points. This analysis also revealed that whether a school is charter has no statistically significant effect on reading scores, b = -.85, t (1,952) = -1.02 , p = .306. The raw regression results can be found below.
The graph below is the regression line, with confidence intervals, showing the negative relationship between student SES status percentage and reading test scores. The red lines represent the average percent of low-income students and the average reading SPG scores in North Carolina schools.
The differences in reading scores between charter and traditional public schools in the state of North Carolina is related directly to the percentage of students in the school living in low-income households and not a result of enrollment in a charter school. This analysis provides further evidence that charter schools do not result in better outcomes for students. Based on this and my previous analysis of the academic outcomes of charter schools not having any significantly better results than traditional public schools, I have to continue to ask:
What is the evidence charter school advocates in North Carolina point to when they make the claim that charter schools are a BETTER alternative?
The evidence does not bear out any conclusion that charters are better. I continue to welcome data to the contrary.