Study Abroad
CASSIE national sample results for study abroad are presented here in an easy-to-digest and share infographic format. Publication and citation information can also be found under ‘Publications and Citation’.
The study abroad analyses results include:
- Effects of study abroad on student success, including graduating in 4 years or 6 years; time to degree; cumulative hours earned; and final GPA.
- Effects of study abroad on sub groups of students, including those that receive need-based aid; underrepresented minority; first generation students; and STEM majors.
- Effects of study abroad program characteristics, including language of instruction; type of program provider; duration of program; and geographic region.
-
Effects of study abroad on student success
Here, the national sample infographic presents descriptive statistics about who studied abroad and who did not, followed by the results of our matching analyses.
The sample consists of the entering cohorts of first-time freshmen from fall 2010 and fall 2011, from 36 institutions; with 30,649 study abroad students out of a total of 221,981 students.
The matching analyses examine the degree to which studying abroad confers an advantage for 5 indicators of student success: graduating in 4 or 6 years, and conditional on graduating - time to degree, final GPA and cumulative hours earned.
The analysis controls or holds constant a slew of academic and demographic factors, which otherwise might be confounded with the decision to study abroad. These include high school GPA, SAT/ACT score, need-based aid receipt, Race/Ethnicity, gender, age at matriculation, major of study, enrollment intensity (full time or part-time), and number of terms enrolled.
-
Effects of study abroad on sub groups of students
Secondly, are the analyses for sub groups of students, those that receive need-based aid, underrepresented minority, first generation students and STEM majors. Please note that for these subgroup analyses, the matching results do not restrict matches to occur within institutions, due to the often low frequency of observations at any single institution. As a consequence, the magnitude of these results are larger than they would be if the institutional restriction was imposed and results should be interpreted with this caveat in mind.
The matching analyses do not capture the causal effect of studying abroad, but rather demonstrate the correlation between studying abroad and student outcomes that remain once characteristics such as pre-college academic preparation (as measured by SAT, etc.) are controlled for. While the CASSIE sample includes institutions from across the United States, the results should not be used to make inferences about practices or expected outcomes at all colleges and universities.
CASSIE results offer rigorous evidence that timely graduation and other elements of student success are in no way impeded by studying abroad. To the contrary, studying abroad boosts students progress to degree. This boost is especially evident for students from subgroups that are typically underserved by education abroad.
-
Effects of study abroad program characteristics
Thirdly, we also conducted analyses which pertain to the effects of education abroad program characteristics on college completion (in 4 and 6 years). The program characteristics we explored are (a) language of instruction (b) type of program provider (c) duration of education abroad, and (d) geographic region in which the education abroad program was conducted. In past analyses, CASSIE has compared students who studied abroad with matched samples of students who never studied abroad. But the objective of this analyses was to identify which education abroad practices confer the greatest benefit on timely college completion. Therefore, records of only education abroad participants were examined. Students who studied abroad in one kind of program were compared with students who studied abroad in different kinds of programs. These analyses are not strictly matching procedures as in past CASSIE work. Rather, they are multivariate regressions which included the same constellation of “control” variables (e.g., age, sex, race, need-based financial aid, etc.). Differences attributable to institutions were also “controlled.” These regressions require us to stipulate one practice as the “reference” or baseline against which other practices are compared. In each instance, we chose the most popular or conventional practice as the reference.