Effects of Personalizing Science of Learning Instruction to Dual Enrollment High School Students’ Prior Knowledge

• In Event: Promoting Cognitive and Contextual Agency Through Scalable Interventions for Self-Regulated Learning and Adaptive Performance

Abstract

Background
States are expanding dual enrollment for high school students (Tobolowski & Allen, 2016) because of the promising results that dual enrollment has shown for high school graduation rates, 4-year college enrollment and persistence (Lee et al., 2022). Despite this initial benefit of dual enrollment, high school students would benefit more from these courses if they were adequately equipped with the self-regulated learning (SRL) skills necessary for college success. High school students struggle with SRL, even more so than their college-age peers (Authors, 2024), making it essential to train them on SRL in an effective and efficient way, before dual enrollment.


Aims
In this study, we tested whether a digital intervention could prepare high school students with SRL knowledge before dual enrollment courses, and if the benefit of this intervention could be enhanced by personalizing it to students' prior SRL knowledge. Personalizing online environments to students’ prior knowledge has been shown to enhance their meaningful learning from these environments (Chen, 2014). Further, this personalization removes extraneous content, thus reducing the time cost and mental effort required by students to complete the intervention. We hypothesized that prior knowledge would moderate the relationship between group (i.e., personalized/non-personalized) and learning outcomes, such that the interaction of higher prior knowledge and personalization would lead to superior learning. We also hypothesized that prior knowledge would moderate the relationship between group and time spent on the intervention, such that the interaction of higher prior knowledge and personalization would decrease time spent.


Method
49 high school students were randomly assigned to one of two groups (personalized, n = 20; non-personalized, n = 29). The personalized group answered formative assessment questions in videos, with feedback, that were aligned to missed questions on the pre-test. The non-personalized group answered all formative assessment questions with feedback regardless of pre-test performance. The intervention included twelve videos teaching the science of learning (Authors, 2023). Students completed a pre-test, then the intervention, in which they watched videos and completed formative assessments. Lastly, students completed the post-test.


Results
A multiple linear regression modelled how pre-test, group, formative assessment performance, and the interaction of pre-test and group predicted post-test. Results indicated the interaction of pre-test and group statistically significantly predicted post-test (B = .951, p = .011). We estimated the same model and replaced post-test with intervention duration as the outcome. The model was statistically significant (p = .04); however, the interaction of pre-test and group did not reach significance (B = -289.19, p = .079). Figure 2 indicates a meaningful interaction; however, the smaller sample size could be limiting power to detect an effect.


Significance
Our findings suggest personalized formative assessments within digital environments, like video-based interventions, is a worthwhile endeavor for teaching students about SRL. Although not statistically significant, our findings indicate higher prior knowledge students with personalized interventions spent nearly 5 minutes less on the intervention, while still experiencing superior outcomes. This expands prior SRL intervention work by testing and finding efficacy for personalized SRL interventions for high school students.

Authors

• Shelbi Laura Kuhlmann, University of Memphis

• C. Noelle Patterson, University of Memphis

• Haley Siegfried, University of Memphis

• Christina Hollander-Blackmon, University of North Carolina - Chapel Hill

• Sarayu Satyasai Yenumula, University of North Carolina - Chapel Hill

• Matthew L. Bernacki, University of North Carolina - Chapel Hill