Introduction Although one-on-one tutoring offers many benefits in terms of flexibility and engagement, it can prove to be a resource-intensive method of instruction. Adaptive training (AT) is a viable alternative when one-on-one tutoring is impractical. AT caters instructional content based on an individual’s aptitudes, learning styles, preferences, or task performance (Landsberg, Van Buskirk, Astwood, Mercado, & Aakre, 2011). The present study describes an experiment performed to examine the effects of different rates of difficulty adaptation to train an audio-visual change detection task. Previous research has shown that difficulty adaptation has benefits for learning (e.g., Landsberg, Mercado, Van Buskirk, Lineberry, & Steinhauser, 2012; Wickens, Hutchins, Carolan, & Cummings, 2013). However, this may not be universally true. Specifically, people who cope with task difficulty using emotion-focused coping (EFC) may exhibit performance decrements when experiencing changes in difficulty. EFC is a method of dealing with tasks or problems involving self-criticism, self-doubt, and worry (Matthews & Campbell, 1998). EFC is also associated with high levels of task distress (Matthews et al., 2006), and previous research has identified that shifts in task workload can impair subsequent performance and increase distress (e.g., Cox-Fuenzalida, 2007; Helton, Shaw, Warm, Matthews, & Hancock, 2008). This presents a problem, as the difficulty adaptations (and their concomitant shifts in workload) that should improve learning gains may be problematic for those who use EFC. Method The present experiment examined the hypothesis that those who use EFC strategies in stressful situations struggle with task performance when experiencing changes in task demands. Ninety-five volunteers completed training in which scenario difficulty adapted based on participant performance. Participants completed five 10-minute scenarios of an audio-visual change detection task. In this task, participants observed and listened to a simulated electronic warfare environment consisting of multiple emitters with unique parameters. These emitters could engage or disengage their signals at any time. Participants were required to submit reports to classify these emitters and indicate when they engaged or disengaged from the environment. They were assigned randomly to one of three groups: within-scenario adaptive where task difficulty changed in real time, between-scenario adaptive where task difficulty changed between scenarios, and a non-adaptive control condition where task difficulty was held constant across scenarios. Data on participants’ pre-training dispositional coping styles were collected using Matthews and Campbell’s (1998) Coping Inventory for Task Stressors (CITS). Additionally, we measured participants’ post-task situational coping styles (CITS), workload (NASA Task Load Index, Hart & Staveland, 1988), and distress (Dundee Stress-States Questionnaire; Matthews et al., 2002) after each of the five scenarios. Results Results indicated no group differences in performance (combined accuracy and timeliness of reports; scored out of 100%), post-task distress, and post-task workload in any of the conditions; however, conditional differences emerged when assessing EFC. We performed a moderated mediation analysis to examine conditional indirect effects of EFC on score, post-task workload, and distress. In this model, dispositional (pre-task) EFC was the predictor variable, situational (post-task) EFC was the mediating variable, and this relationship was moderated by the training condition. The outcome variables were score, post-task distress, and post-task workload. This analysis was performed on all five scenarios, and revealed significant indirect effects for those high in dispositional EFC, such that their performance was worst and their distress and workload were highest in the within-adaptive condition. In the between-adaptive and control conditions, these effects were present in the first scenario but diminished over time. In essence, real-time difficulty adaptation that is too frequent can lead to poor training outcomes for those high in EFC. Discussion These effects are consistent with those of Cox-Fuenzalida (2007) and Helton and colleagues (2008) who suggested that variable shifts in workload could be detrimental to performance. Our results suggest that those who tend to use EFC also tend to struggle with frequent changes in task difficulty. For training purposes, EFC may be an important variable for adaptation prior to training (i.e., macro-adaptation; Park & Lee, 2003). If an adaptive training system offers multiple schedules of difficulty adaptation frequency, one could assign higher-EFC trainees to a less-frequent difficulty adaptation schedule to minimize the risk of lower performance and increased distress and workload. On a similar note, it may also be beneficial to take periodic measurements of trainees’ situational EFC to shift their adaptation schedule to one that is less frequent. In both cases, these approaches to adaptive training would mitigate the negative training outcomes of frequent difficulty adaptation associated with high EFC. Future research should continue to study how trainee individual differences influence the effectiveness of AT. Acknowledgments We gratefully acknowledge Dr. Kip Krebs and the Office of Naval Research who sponsored this work (Funding Doc# N0001417WX00200). We would also like to thank Mr. Derek Tolley for his development of the experimental task. Presentation of this material does not constitute or imply its endorsement, recommendation, or favoring by the U.S. Navy or Department of Defense (DoD). The opinions of the authors expressed herein do not necessarily state or reflect those of the U.S. Navy or DoD. NAWCTSD Public Release 19-ORL045 Distribution Statement A – Approved for public release; distribution is unlimited.
Adaptation supports short-term memory in a visual change detection task
