As part of an effort to provide evidence for the reproducibility of educational studies for a variety of student bodies, we collected a year-long data set in introductory physics courses at Cal Poly Pomona (a primarily undergraduate and Hispanic-serving institution) to understand factors that affect students' performance on the FCI (force concept inventory). This study also allows us to gain insights into possible gender or racial gaps in students' performance. In this paper, we discuss background variables that predict students' FCI scores at the end of the term. Weak correlation is found between students' SAT score and FCI normalized gain, suggesting education research findings can be population dependent. Gender and racial gaps are found in students' FCI performance, both at initial preparation and overall gain. There is a gender gap of 16% FCI pre-test and 17% of FCI post-test with a strong effect size (d=0.81). Caucasian students outperform Asian and Hispanic students on FCI pre, post and gain. No significant interaction is found between gender and race/ethnicity after controlling for course grade.