Both the course and the assistant increase behavioral engagement, measured as characters written per journaling prompt¶

Claim · All claims
Evidence · 1 study · causal r2 · q3
Currency · newest study 2026 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q3 i? In a multiple regression on message length, both feature access and study day significantly predicted the number of characters written, confirming H3 and H4. → Sven Scheu 2026

Evidence¶

Sven Scheu 2026¶

Sven Scheu, Simone N. Loeffler and Alexander Maedche. (2026). Designing a mobile chatbot-based learning journaling system for intrinsic motivation and engagement. International Journal of Educational Technology in Higher Education. https://doi.org/10.1186/s41239-026-00589-7

q3 · i? · causal · r2

Multiple regression modeling characters written per prompt (5181 unique messages after outlier removal; R²adj = 0.06, F(4, 1616) = 24.84, p < 0.001) in the field experiment. Table 2 prints unstandardized coefficients of Assistant B = 18.32 and Course B = 22.95 (both p < 0.001) and a negative Course * Assistant interaction of −15.21; mean message lengths were longer in all treatment groups (A: 71.25, C: 79.26, CA: 83.04) than baseline (51.21).

"the regression shows that both features had a significant impact on the amount written by the users, thus confirming H3 and H4."

Discussion¶

Related Claims¶

• AI-assisted journal entries show higher lexical diversity and more emotion-word usage than unassisted entries — related
• Question-driven AI journaling increases intrinsic motivation and acting confidence among actors — related
• The example-based course increased perceived competence, but neither the course nor the assistant had a significant effect on perceived choice or pressure — related
• The LLM-based journaling assistant shows no significant effect on students' intrinsic motivation for creating a learning journal — related
• No significant medium, context, or medium-by-context effects on written response length as a measure of behavioral engagement — reports the opposite
• Reading comprehension uniquely adjusted behavioral engagement, with better comprehenders investing more time and writing longer responses — related
• AI-generated questions instead of draft text significantly reduce actors' cognitive burden during character journaling — related
• Course users create their journal entries earlier relative to the daily notification, while assistant access shows no significant timing effect — related