ORCID

https://orcid.org/0000-0001-5725-6734

Date of Award

Fall 2025

Language

English

Embargo Period

12-1-2025

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

College/School/Department

Department of Educational and Counseling Psychology

Program

Educational Psychology

First Advisor

Kimberly Colvin

Committee Members

Heidi Andrade, Angela Lui, Jason Bryer

Keywords

Higher education, college readiness, adult learners, item response theory (IRT), multistage adaptive testing (MST), differential item functioning (DIF)

Subject Categories

Adult and Continuing Education

Abstract

This dissertation examines age-based differences in college readiness by comparing newly enrolled adult undergraduates (age 24 and older) with traditional-age undergraduates at two public institutions: a large online university serving primarily working adults and a residential research university serving primarily younger students. The study uses secondary data from 5,447 incoming students who completed the Diagnostic Assessment and Achievement of College Skills (DAACS) Reading and Mathematics assessments between May 2022 and May 2023. The primary goal was to test whether adult and traditional-age students differ in reading and mathematics readiness after adjusting for background characteristics.

Multistage bifactor and unidimensional two-parameter logistic item response theory models were used to evaluate internal structure and item quality. Two item pools (180 reading and 174 mathematics items) were tested for differential item functioning (DIF) by age group, gender, race/ethnicity, socioeconomic status, military affiliation, and transfer-credit status using logistic regression with IRT-based ability estimates. To prioritize measurement invariance, a deliberately conservative DIF-flagging strategy was adopted, and low-performing items were excluded; total reading and mathematics scores were then re-estimated from the refined item pools. Multiple linear regression models, supplemented by sensitivity analyses, tested age differences in academic readiness while controlling for the same set of covariates.

After covariate adjustment, reading readiness did not differ meaningfully by age, whereas adult students showed significantly lower mathematics readiness. These results underscore the need to examine differences between the more frequently studied population of traditional-age students and the understudied, rapidly growing population of adult learners in order to equitably accommodate both groups.

License

Creative Commons Attribution 4.0 International License
This work is licensed under a Creative Commons Attribution 4.0 International License.

Download

Share

COinS