Examining Pathways: Do High-Stakes Tests Predict Postsecondary Completion?

• In Event: Excellence in Education Research: Early-Career Scholars and Their Work

Abstract

With the rise of the educational accountability movement, research has demonstrated that using high stakes tests (HST) as measures of achievement is highly problematic and has a negative impact on students’ educational journeys (Vasquez Heilig & Darling-Hammond, 2008; McNeil, 2005; Figlio & Getzer, 2002). In particular, HST have the most impact on ethnic/racial and linguistic minority students, students with disabilities, and students who live in low-income areas (Menken, 2010; Garcia & Gopal, 2003). With the increase in HST, such as high school exit exams, there has been an increase in dropout rates and these exams have been identified as a contributing factor in the school-to-prison pipeline (Thompson & Allen, 2012). While extensive research has been conducted in the previous areas, the relationship between high stakes testing and postsecondary completion remains understudied.

Building on previous research (Rodríguez & Arellano, 2016), this study incorporates multiple sources of data along the educational pathway. Math and English test scores from students’ 8th grade are followed up with California High School Exit Exam (CAHSEE) scores and deeper learning constructs at the high school level are included, high school completion, followed by postsecondary enrollment and degree attainment data acquired from the National Student Clearinghouse, culminating in institutional-level data from the Integrated Postsecondary Education Data System (IPEDS). Each of these longitudinal data points provide for a rich dataset to understand the educational pathways of 8,279 students from the state of California.

The guiding research questions for this study are: What are the educational trajectories of students with varying Math and English test scores? Do deeper learning factors predict postsecondary success? What kind of institutional environments (i.e. college and university characteristics) ensure degree attainment?

A total of five school districts in the state of California are included in the dataset. Women make up 52% of the sample (n = 4,330). The sample also includes a diversity of students from the various racial/ethnic demographic categories: Black (n = 1,010), Hispanic (n = 3,217), White (n = 2,668), Asian (n = 1,125), and Other (n = 222). Enrollment patterns demonstrate that the 8,279 students in the dataset enrolled in and/or graduated from a total of 609 unique colleges or universities. A total of 1,078 students graduated with a certificate, 2-year, or 4-year degree.

Student background characteristics, high school information, National Student Clearinghouse data and postsecondary information are combined to form the dataset. Descriptive statistics were conducted to understand the context of the multiple educational environments. Correlations also demonstrated significant relationships between the variables. Finally, a hierarchical generalized linear model reveals the unique influence of student-level variables alongside institutional-level variables on the outcome of postsecondary degree completion.

Authors

• Lucy Arellano, Oregon State University

• Jessica Melinda Rodriguez