Return to search

Using Early Alert Data to Increase Success Rates Among U.S.

More U.S. students are attending college than at any time in history, but many of these students are poorly prepared for college coursework. Higher education institutions are challenged to increase the overall student success rate. The study community college implemented an early alert system to identify students with potential course performance concerns, but it is not known whether other student data might also predict academic performance at the study site. Guided by Tinto's and Astin's respective works on student persistence, the purpose of this correlational study was to investigate the relationships between students' demographic, background, and environmental variables and course success, for students identified by the early alert system. Stratified random sampling of 4 academic years of student data using 50% of the early alert students in the top 25 courses with the highest number of early alert students yielded a sample of 3,873 students. Predictor variables were gender, race/ethnicity, age, income status, campus, faculty status, first generation to college, and course times. The dependent variable was the course outcome. A regression analysis examined the predictive ability of each variable, and race/ethnicity was the only predictor found to be statistically significant. African-American students were the highest risk students for failing a course. A white paper was developed to share the study findings with the administration at the study college regarding the early alert system and other factors related to course success. Increasing student success may produce an overall positive social impact on society by increasing students' job prospects and ability to contribute economically in their communities.

Identiferoai:union.ndltd.org:waldenu.edu/oai:scholarworks.waldenu.edu:dissertations-4241
Date01 January 2016
CreatorsBall, Donald Michael
PublisherScholarWorks
Source SetsWalden University
LanguageEnglish
Detected LanguageEnglish
Typetext
Formatapplication/pdf
SourceWalden Dissertations and Doctoral Studies

Page generated in 0.001 seconds