ComBi

Digital tools have the potential to facilitate learning by:

• reducing cognitive load (Skulmowski & Xu, 2021),
• bolster long-term learning motivation (McGuire et al., 2022), and
• encouraging open-ended, inquiry learning (Griffith et al., 2024).

But when learning is too open-ended, it has poor outcomes for students who doubt their own ability or effectiveness (i.e., have low self-concept) in the subject (Richter et al., 2022).

Few studies have tested how self-concept impacts inquiry learning in applied math biology.

Learning math requires more self-study (Spitzer 2021), traditionally via pen-and-paper textbook exercises. To see if low self-concept limits the effectiveness of inquiry learning, we compare an interactive web-app learning task (top left in figure) with both video demonstration (top right in figure) and textbook learning (bottom left in figure) tasks.

Our population of interest is ≥2nd year biology bachelor students from universities with some English-language instruction. The online learning task takes 30 minutes. Participants are randomly assigned to one of three learning tasks (see figure above) and case studies (see section below). Each case study uses dynamical systems with alternative stable states and bistability. This enables us to assess a common set of learning outcomes:

1. The long-term outcome of a system depends on how many attractors it has.
2. The parameter values control whether there is a single attractor or multiple.
3. With multiple attractors, the long-term outcome depends on the initial conditions.

By comparing participants’ responses to case study-specific scientific questions before and after the learning task, we can identify the type of learning task that best supports different students in becoming better mathematical modelers.

For more details about our methods, check out our study preregistration at this link.