EA infographic 2021

KSU Education Abroad Graduation and Retention Data

EA Graduation and Retention

Analysis of Education Abroad Participant GPA, Retention, and Graduation Outcomes

The Office of Institutional Research and Assessment and the Education Abroad Office have partnered to analyze the impact education abroad participation has had on retention and graduation outcomes for KSU students that have participated in an education abroad program.

As a high-impact practice, education abroad’s benefits on student outcomes are well established in research literature. This study attempts to identify whether these broader research conclusions hold true among education abroad participants at Kansas State University.

Student outcomes analyzed included latest GPA, four-year graduation rate and six-year graduation rate. For the GPA analysis, only degree-seeking undergraduate students who studied in K-State between 2012 and 2020 and have a GPA were included; For four-year graduation rates, only degree-seeking undergraduate students who entered K-State in or before 2016 were included; For six-year graduation rates, only degree-seeking undergraduate students who entered K-State in or before 2014 were included.

The significance of differences in the latest GPA was examined using a t-test, while the significance of differences in graduation rates was examined using a Chi-sq test.

Results and Analysis

A comparison without matching was conducted first. In terms of the latest GPA, four-year graduation rate and six-year graduation rate, participants in an education abroad program performed better than non-participants, and the differences are all significant.

Comparison with Unmatched Cohorts

Mean latest GPA

Four-year graduation rate

Six-year graduation rate

Non EA-Participants

3.07

47.47%

64.68%

EA Participants

3.41

56.66%

92.79%

Significance (Chi-sq/t)

<.0001

<.0001

<.0001

However, a comparison like this may not reveal the actual performance gap between participants and non-participants because there might be a pattern in education abroad program participation. For example, students with a better high school GPA may be more likely to participate in education abroad programs. To better compare similar groups of students, this study used propensity score matching to establish cohorts of participants and non-participants with the same or similar features, and then compared their performance with matched cohorts. In the matching process, exact matching was conducted based on race/ethnicity, gender, high school GPA and college. The race/ethnicity group was composed of a non-underrepresented minority group including white students and Asian students and an underrepresented minority group including students of other races. High school GPAs were divided into five intervals based on percentiles.

Over 90% of participants have at least one matched non-participant, so exact matching was used for this study.

Sample Size with and without Matching

Mean latest GPA

Four-year graduation rate

Six-year graduation rate

Non-Matched

Non-EA Participants

47499

60603

50815

EA Participants

4394

4128

2982

Matched

Non-EA Participants

12337

11658

8622

EA Participants

4361

4087

2982

Through comparisons with matched cohorts, participants of education abroad program still perform significantly better than non-participants in latest GPA, four-year graduation rate and six-year graduation rate. But the gaps between the two groups are a little narrower than the gaps without matching.

Comparison with Matched Cohorts

Mean latest GPA

Four-year graduation rate

Six-year graduation rate

Non-Participants

3.19

49.99%

66.46%

Participants

3.41

57.08%

92.79%

Significance (Chi-sq/t)

<.0001

<.0001

<.0001

Conclusions

Participants in education abroad perform better across several academic outcomes compared with non-participants both when comparing all students and when comparing students with similar preparedness and demographic variables. When compared with non-participants with the same characteristics, participants’ advantage is a little smaller than being compared with all non-participants.

More factors in addition to race, gender, high school GPA and college can be included in the matching process to further improve similarity of the two cohorts being compared.