IN APRIL 2006, this column summarized an analysis of data from the National Assessment of Educational Progress by Christopher Lubienski and Sarah Theule Lubienski indicating that most of the differences between test scores in public and private schools could be accounted for by demographic differences between the two types of schools. The publics had more poor students, more minorities, more special education students, and more English-language learners. That study--and others with similar findings--was criticized in some quarters because it was a cross-sectional analysis, a picture snapped at a single point in time. Now comes Harold Wenglinsky of Columbia University with an analysis commissioned by the Center on Education Policy that looks at the same students in grades 8, 10, and 12 and at age 26. The results are similar to those of the cross-sectional studies. Wenglinsky's study also went beyond the public/private dichotomy that oversimplifies the situation. In the new study, he examined public high schools, magnet schools, public schools of choice, Catholic diocesan schools, Catholic holy order schools, Catholic parish schools, non-Catholic religious schools, and independent private schools. Of course, kids don't enter high school from a vacuum. Fortunately, in this sample taken from the National Educational Longitudinal Study, eighth-grade test scores are also available. Finally, kids don't live in a vacuum outside of schools. This study looked at socioeconomic status (SES), parental expectations, parental discussions about schoolwork with the children, and parent involvement in school activities. Analyses were conducted on reading, mathematics, and history achievement in the 12th grade. For reading, all three parental variables were positively related to 12th-grade achievement, as was SES, but the strongest relationship was between eighth-grade and 12th-grade test scores. Only the religious order schools showed a positive relationship with the parental variables, and the relationship was much smaller than that of SES. For math, parental discussions about schoolwork and parent involvement in school activities had no on achievement, although SES and parental expectations did. Again, only the Catholic religious order schools outperformed public schools. For science, nothing other than SES had any impact. No type of school showed any advantage. Parental expectations and discussions showed up again as important in history achievement, along with eighth-grade test scores. Once again, Catholic religious order schools were the only type of school to show any advantage over publics. At age 26, students who attended private schools were no more likely to be civic-minded or satisfied with their jobs than those from public schools. Civic-mindedness included measures of whether students had participated in a youth organization, a civic organization, or a political campaign or whether they had voted in the most recent election. They were asked about job satisfaction in terms of pay, fringe benefits, challenge of the work, opportunities for advancement, job security, and opportunities for further education and training. The strength of Wenglinsky's study is that it is longitudinal, running from eighth grade through 12th grade and picking up some nonacademic variables when the students were 26 years of age. Most other studies of public/private differences have been cross-sectional. The Fordham Foundation criticized this study because it studied only students who had completed high school and because the entire sample came from the bottom quartile of family income. These were both deliberate choices. The study wanted to see the full impact of high school and thus needed graduates. It would have been nice, though, if the study had included some indication of attrition. Do the privates hold their students better, or do they just say, so long it's been good to know you? …
Read full abstract