Graduation Year


Document Type




Degree Name

Doctor of Philosophy (Ph.D.)

Degree Granting Department

Educational Leadership

Major Professor

Victor M. Hernández-Gantes, Ph.D.

Co-Major Professor

Yi-Hsin Chen, Ph.D.

Committee Member

William Blank, Ph.D.

Committee Member

Judith A. Ponticell, Ph.D.


STEM students, demographic characteristics, academic performance, institutional characteristics, logistic regression, workforce


Today, economic growth, technology advancement, and daily life are all driven by science, technology, engineering and math (STEM). American colleges and universities need to produce more STEM graduates in order to meet rapid job growth. The purpose of this study was to determine the significant factors that influence the persistence of STEM students in community colleges. The research described the characteristics of STEM students who persisted and of those who did not in STEM, as well as the different characteristics between male and female students. It also investigated student and institutional factors that influenced STEM students’ persistence. Furthermore, the study examined the persistence of male and female students as two separate groups and found out the differences in the influence of various factors on male and female STEM students in community colleges.

The research design of the study is a quantitative, causal-comparative, and an ex post facto design. Data collection included student-related variables such as demographic background, academic performance, and education experience. Institutional variables covered institution fall enrollment, degree of urbanization, and percentage of white enrollment. All the data were sourced from the U.S. Department of Education’s National Center for Education Statistics’ (NCES) Beginning Postsecondary Longitudinal Study:12/14 (BPS:12/14). PowerStats, a web-based tool from NCES, was used to analyze the data by descriptive and logistic regression methods.

The descriptive analysis showed that higher proportions of STEM persisters demonstrated the following background and demographic characteristics: female, age below 18, White, parent’s highest education level with some college experience, Federal Pell Grant recipient, and with none or one index of risk factors. In terms of academic and college experience, the persisters were more likely to have an SAT score between 1000 to 1150, did not take any remedial courses, strongly agree on academic confidence, with the highest education level expected bachelor’s degree or above, accumulated college GPA above 2.75, did not quite feel like a part of the institution, strongly agree on engagement with faculties, and did use some PSE services.

The logistic regression demonstrated that three factors played significant roles in STEM students' persistence in community colleges, which were Federal Pell Grant received years, parents’ highest education level, and tuition and fees paid. Results showed that the more years that students received Federal Pell Grant, they were more likely to persist; students with parent’s highest education level above bachelors’ degree were more likely to persist, and tuition and fees paid did not have much effect on students’ persistence. Due to the small sample size for male and female STEM students, no significant factors were generated in final logistic regression models.

This study contributed to the literature on factors that influence STEM students’ persistence in community colleges.