Open educational resources and student performance trajectories: B is achievable, A illusive

ABSTRACT

Open Educational Resources (OER) are regularly adopted, yet questions persist about efficacy. Do OER improve student performance? Research produces ambiguous results, because these are derived from small sample sizes, inadequate efficacy measures, and neglect of interactional effects. We examine the effect of OER on student grades using over 500,000 courses, including 72,000 students during four years (2017-2020), from a community college in the Northeast of the USA. Relying on logistic regression models, we test the probability of passing and passing with the highest grade. Students who participate in OER initiatives have higher odds of both. Concurrently, these odds are affected by demographic and academic factors. The odds of passing are higher in OER classes for students in good academic standing, entry level, face-to-face classes. The odds of passing with a high grade are more nuanced. Minoritized students in advanced level, in-person OER courses are more likely to earn a “B.” White students in entry level, online, non-OER courses are more likely to earn an “A.” Overall, OER help most students pass up to a “B.” However, they fall short of helping select students reach an “A.” If OER are to replace traditional textbooks, more efforts are needed for maximal productivity.

KEYWORDS:

• Open educational materials
• academic success
• student success

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

1. GPA stands for grade point average, an indicator of student standing. DFW stands for letter grades; D is a low passing grade, F means the student failed the class, W means the student withdrew from class.

2. With over 70,000 students over a four-year period taking an average of 7–8 courses per year, we end up with more than 500,000 student courses as our sample size. Refer to Tables 1 and 2 for clearer student-course distributions.

3. The lowest category (designated as ‘1’) in the independent and control variables is used as a reference group in logistic regression models.

4. We opted for a single unit regression model, in part, due to the computational complexity of mixed models. With mixed models, interaction effects make the calculations difficult, if not impossible. We have quite a few variables that need to be included in the models and, more significantly, we have interaction effects that we think are the essence of this analysis.

5. Note, in our report, significance is less than .001, unless otherwise specified.

6. As a reminder, the probability of failing is the opposite of the probability of passing. Our discussion centres on passing only.

7. The probability of passing also depends on cumulative GPA. However, the differences are minimal, and, for parsimonious reasons, we do not present them in a graph. If needed, the graph is available upon request.

8. For parsimony, we only discuss the probability of earning an ‘A’ or a ‘B.’ Probability tables for ‘C’ and ‘D’ grades are available upon request.

Additional information

Funding

The work was supported by the City University of New York and the New York State Education Department

Notes on contributors

Ilir Disha

Ilir Disha is an Associate Professor in Criminal Justice at the Borough of Manhattan Community College (BMCC). He has various research interests. In addition to a recent focus on Open Educational Resources, his work has explored the role of triggering historical events on hate crimes, immigrant assimilation on neighbourhood crime, and student experiences of victimization or police encounters on choice of major and career motives. Currently, he is working on a couple of new projects; one that investigates the recent political climate on hate-motivated behaviour and another that examines the outcomes of the First Step Act policies on federal prison reform.

Brenda K. Vollman

Brenda K. Vollman is an Associate Professor in Criminal Justice at the Borough of Manhattan Community College (BMCC), City University of New York (CUNY). In addition to a recent focus on Open Educational Resources, her research interests are oriented around understanding the nexus of victim, offender, and criminal event context. This is explored using data derived primarily from written accounts and interviews in order to understand the ways in which we perceive and construct narratives of violence and victimization (particularly sexual victimization), and how these shape social and institutional responses.