Factors Associated with High School Exit Exam Outcomes Among Homeless High School Students

Schedule:
Saturday, January 17, 2015: 10:00 AM
Preservation Hall Studio 1, Second Floor (New Orleans Marriott)
* noted as presenting author
Mathew C. Uretsky, MSW/MPH, Student, University of Maryland at Baltimore, Baltimore, MD
Susan Stone, PhD, Associate Professor, University of California, Berkeley, Berkeley, CA
Background and Purpose

Homeless elementary and middle school students are at risk for poor academic outcomes, including low grade point average and standardized achievement scores, high frequencies of absenteeism, and suspension. In contrast, little is known about how the experience of homelessness affects high school student achievement. This study begins to address current gaps in research by examining factors related to homeless students’ high school exit exam completion and performance. Exit exam performance is ideal high school outcome of interest as it measures the minimum skill level required to receive a high school diploma. Specifically, this study identifies key demographic and academic performance factors that affect the odds of attempting and passing the California High School Exit Exam (CAHSEE) for students identified as homeless by a district’s McKinney-Vento program.

Methods

Student-level data were obtained from a large urban district in California. The sample included all homeless-identified students eligible to take the CAHSEE (10th-12thgraders; n=416). The dependent variables were dichotomous indicators indicating (1) whether a student took the CAHSEE (versus not) and (2) whether a student passed the CAHSEE. Independent variables included days absent, times suspended, and cumulative GPA as well as dichotomous indicators reflecting gender, ethnicity, English fluency, grade level and performance on standardized language arts and mathematics tests (below basic or score missing versus above basic).  Models were estimated using binary logistic regressions with robust clustered standard error to correct for student clustering within schools.

Results

Being in 11th or 12th grade (compared to 10th), having a lower GPA and having no recorded score for the standardized mathematics test significantly increased the odds that student would not attempt the CAHSEE tests. The overall model accounted for 12% of the variance in the dependent variable and correctly classified 69% of the cases.

English fluency and belonging to the Latino or “other” ethnicity (compared to African-American) were associated with decreased odds of passing both CAHSEE tests. In addition, scoring below basic or having no recorded score on the standardized language arts and mathematics tests (compared to above basic), and having a lower cumulative GPA were also associated with decreased odds of passing the CAHSEE. The overall model accounted for 32% of the variance in the dependent variable and correctly classified 79% of the cases.

Conclusions and Implications

Past test taking behavior, grade level and cumulative GPA relate to whether a student attempted the CAHSEE. For students who do attempt the CAHSEE, factors that have been used to predict graduation, such as attendance and suspension, are not related to passing. Instead, passing the CAHSEE appears to relate to development of basic skills (Cumulative GPA, STAR ELA scores), suggesting that indicators used for identifying risk of non-graduation in earlier grades may not be appropriate at the high school level, at least, among McKinney-Vento identified youth.  Current programing designed for preventing skill loss during an episode of homelessness may not be sufficient to foster academic success among high school students experiencing homelessness.