Video Modeling for Training Older Adults to Use New Technologies
Doreen Struve, Hartmut Wandke · 2009 · ACM Transactions on Accessible Computing · doi:10.1145/1525840.1525844
Summary
This paper investigates the effectiveness of video modeling combined with guided error training for teaching older adults to use complex technology systems. Part of the ALISA project (Adaptive Learning Support for Older Adults) at Humboldt University Berlin, the study compared 40 younger adults (mean age 24) with 40 older adults (mean age 67) learning to use a simulated Ticket Vending Machine (TVM) for Berlin public transport. The training program incorporated multiple instructional design principles optimized for older learners: slow pacing (8 seconds per interaction step), short video segments (averaging 1.2 minutes), simple vocabulary (Flesch readability index 64), high-contrast displays, verbal markers instead of icons, and a "training wheels" approach that progressively introduced complexity by graying out irrelevant functions. The video models were older adults themselves, addressing socio-motivational aspects by demonstrating that people like the learners could successfully use the technology. The key experimental manipulation compared guided error training—where the model demonstrated typical errors and their corrections—against error-free training where only correct procedures were shown. The study grounded this approach in Bandura's Social Cognitive Theory and Mayer's cognitive theory of multimedia learning, arguing that dual encoding (visual + auditory) optimizes learning while demonstrating errors helps learners build richer mental representations of system structure.
Key findings
Both younger and older participants who received guided error training made significantly fewer errors in knowledge transfer tasks compared to error-free training (F(1,76)=11.85, p<.001, η²=.14). Guided error training also produced significantly greater gains in structural knowledge—understanding of how the TVM system is organized—particularly for older adults (F(1,38)=7.12, p<.01, η²=.16). Self-efficacy beliefs increased significantly for both age groups after training (F(1,76)=29.77, p<.001, η²=.28), though the training condition did not differentially affect self-efficacy gains. Importantly, presenting errors during instruction did not increase perceived cognitive load; participants in the guided error condition actually rated navigational demands as lower than those in the error-free condition (F(1,76)=6.90, p=.01). Age differences persisted across measures: older adults made more errors, took longer, and used more additional steps than younger adults in both conditions. However, guided error training was particularly beneficial for older adults' structural knowledge gains, suggesting that seeing errors and corrections helps build better mental models of complex systems.
Relevance
This research provides evidence-based guidance for designing accessible technology training for older adults—a growing user population often excluded from digital services due to inadequate support. The specific instructional design recommendations (slow pacing, progressive complexity, high contrast, older adult models, verbal over iconic labels) offer a practical framework for creating inclusive training materials. The finding that guided error training improves learning without increasing cognitive load challenges the assumption that training should always present only correct procedures. For accessibility practitioners designing onboarding flows, help systems, or training materials, showing common errors and their corrections may help users build more robust mental models and recover more effectively when problems occur. The study's limitations include a highly educated sample and small effect sizes, suggesting the training program's baseline accessibility features may have already addressed many barriers. Future research should examine whether guided error training benefits extend to other domains like healthcare technology, where error prevention is critical and older adult users are common.
Tags: older adults · video modeling · training · self-efficacy · cognitive load · guided error training · instructional design