Sunday, March 29, 2009

School Choice in Hoboken- A Statistical Analysis

This data was previously presented at a Board of Education meeting in the Fall. Due to a number of requests, I am publishing the statistical analysis here. To be clear, this analysis simply looks at whether there was any statistically significant differences between Choice children and Resident children on 3 different dimensions. All students were enrolled in Hoboken High School from 2004-2007 (latest data available at time of the analysis). -Dr. Petrosino

A quantitative analysis was conducted on existing student data to see whether there was any significant difference between two populations of students. Population One included 27 graduating seniors from Hoboken High School during academic years 2004-2005 (n=2); 2005-2006 (n=7); and 2006-2007 (n=18) who were designated SCHOOL CHOICE. Population Two consisted of a randomly chosen matching subset of students (matched on gender and graduation year) enrolled in Hoboken High School as city residents (HHS). Therefore, we essentially took our known high school universe of school choice students (SC) and matched them with a comparison group of non-school choice high school students from the same cohort (HHS).

These two groups were then compared across three dimensions. Dimension 1 consisted of Attendance/Behavioral measure (Absences and Lateness). Not only are these figures indicative of behavior but also collate well with behavioral data such as detention/suspension (which by law we do not have access to upon graduation). Dimension 2 consisted of class grades. We chose the last English and last Mathematics course taken by each student. Finally, Dimension 3 consisted of standardized test measures. In this case, we choose to look at the SAT Math, Writing, and Essay scores.

Results:
Dimension 1: School choice children were absent 2 less days a year (13.29 vs 15.4) and were late ½ day more (11.85 vs. 11.22). Absence was statistically significant (.05) indicating that SC children were absent less often but difference in lateness was not statistically significant.

Dimension 2: Class grades in English (82.7 vs 82.09) and Mathematics (78.5 vs 78.22) were not statistically different and indicate no difference between the two groups.

Dimension 3: Non School Choice (HHS) children scored higher in SAT Math (435 vs 409) and SAT Writing (404 vs. 379) by roughly 25 points while School Choice (SC) students outperform their non School Choice peers on the SAT Essay (6.38 vs 5.6). While some minor differences  exist, there is no clear pattern that seems to favor either group.

Conclusion: Based on the data to date, there appears to be little to no statistically significant difference between the two student populations on any of the three dimensions of Attendance/Late; Grades; Standardized Test Scores.

DATA
Abs Late EngGrade MathGrade SATMATH SATWRIT SATESSY
SC 13.29 11.85 82.7 78.5 409 379 6.38
NC 15.4 11.22 82.09 78.22 435 404 5.6

Picture: (L-R) "Sonny" Palmeri, Steve Cappiello, and Mayor Louis DePascale circa Fall 1968. Petrosino Club, Fourth and Monroe St. 

1 comment:

marry said...

oho good dear !!!! very interesting blog and a good posting !!! you must maintain your blog, its interesting !!! Nice Buddy
________________________________

Buy A Research Paper