• Calculus is a gateway — and a barrier — to STEM degrees (Redmond-Sanogo et al., 2016).

• Half of all undergraduates attend community colleges (CCs), but are less likely to earn STEM degrees (Bahr et al., 2023; Zhang, 2022).

• CCs have reformed math pathways: corequisite courses, concurrent/dual enrollment, and moves toward holistic placement.

• But we lack evidence on whether these reforms predict calculus success — especially for students who are underrepresented racially minoritized (URM) (Bahr et al., 2017; Burn et al., 2015).

RQ1 — Preparation, Placement, & Pathways

To what extent does HS GPA and ACT Math predict calculus success, and how are these moderated by time since high school or concurrent enrollment?

RQ2 — Student Characteristics & Learning Contexts

To what extent do demographics and prior math coursework predict success, and how are these moderated by corequisite course availability or section mean HS GPA?

Note. HS GPA = high school grade point average, ACT Math = ACT Math Score, concurrent enrollment = student took concurrent enrollment prerequisite prior to calculus, demographics = URM and biological sex, prior math coursework = the number of CC math prerequisites prior to calculus.

• Theoretical Framework - Academic momentum conceptualizes continued academic achievement as a result of past course taking behaviors and academic achievements (Adelman, 2006). This framework has been extended to momentum for CC student success (Wang, 2017) and STEM academic momentum (Wang, 2015; Zhang, 2022).
• Multilevel Logistic Regression - students (level 1) nested within course sections (level 2) (McCoach & Cintron, 2021; Raudenbush & Bryk, 2001).
• Wyoming CC System - 5 of 7 colleges, years 2013-2023.
• Analytical Sample - \(n=745\) students in \(k=91\) sections (after exclusions).
• Outcome Variable - Binary, Pass (A,B,C,S) or Fail (D,F,U) in Calculus I.
• 13 Predictors - 7 student-level, 2 section-level, 2 student-level interactions, 2 cross-level interactions.

Note. The intercept \(\gamma_{00}\) represents the grand mean log-odds of passing, and the random effect \(u_{0j}\) captures section-specific deviations from this grand mean. \(\gamma\) coefficients represent the expected change in log-odds of passing CC calculus for a (unit) change in the variable. Level 1 = student, level 2 = section. Grand mean centered = gmc, centered within sections = cwc. Bold coefficients address RQ1.

Note. The intercept \(\gamma_{00}\) represents the grand mean log-odds of passing, and the random effect \(u_{0j}\) captures section-specific deviations from this grand mean. \(\gamma\) coefficients represent the expected change in log-odds of passing CC calculus for a (unit) change in the variable. Level 1 = student, level 2 = section. Grand mean centered = gmc, centered within sections = cwc. Bold coefficients address RQ2.

The reforms are real, but the tracking logic persists — inequitable placement policies and who benefits from peer composition and concurrent enrollment.

• Data: Administrative records — no direct measures of placement processes, student aspirations, or instructional quality.
• Sample: Rural Wyoming CC system — generalizability is limited. Much missing data.
• Unmeasured factors: Students’ secondary mathematics history, socioeconomic context, geospatial access (Sonnert et al., 2016).
• Future work: Qualitative investigation of URM student experience in calculus sections, study of corequisite implementation fidelity (Stone-Johnstone, 2023), research Wyoming dual/concurrent enrollment calculus students (Corin et al., 2020), and equity in dual enrollment course access (Xu et al., 2021) .

Reforming the Math Hierarchy: Policy, Practice, and the Politics of Equity in Access and Pathways. AERA 2026.

Breaking or Building Tracks? Community College Mathematics Access and Calculus Success

Michael Bostick | bostickmike@gmail.com | Central Wyoming College

Full model results, diagnostics, and summary statistics:

Multilevel Logistic Regression Model Dashboard