Honors Thesis Research, HCI/ Distributed Cognition Lab, UC San Diego
La Jolla, CA
Under the guidance of faculty Professor Jim Hollan and Professor Scott Klemmer and PhD student Adam Rule, I independently researched what review methods work best in the context of working tasks and online learning. This independent undergraduate research program required defining a question to be investigated, survey existing literature, describing the approach and methods used, explaining how data will be collected, recruiting human requirements and discussing expected results.
The question of how to minimize the impact of interruptions and reestablish the context of interrupted activities remains an open one. Systems that restore work context must be concerned not only with restoring computing resources, but also with helping users remember forgotten motivations and methods. However, little is know about what makes for a good memory cue or what types of memory are brought to mind by particular types of cues. These studies begin to address this knowledge gap by presenting two exploratory studies of visually-cued memory for computing tasks.
"That Reminds Me: Identifying Elements of Screen Recording that Cue Contextual Memory"
In experiment one, PhD candidate Adam Rule and I analyzed cue-memory pairs drawn from sessions of reviewing screen recordings and make two main contributions. First, our paper clarifies which visual elements cue memory, specifically finding that new windows, content editing and static text account for most of the observed cues. Secondly, its explores what types of memory are prompted by viewing past computer activity, finding that memories of What happened are most common, followed by memories of How and Why an activity was taking place.
"Learning with Interruptions: Representing Past Computer Activity for Recall of Learned Procedures"
I am interested in how people revisit previously begun work tasks. Furthermore, what review systems prompt the most relevant contextual memory. I conducted an experiment where participants were asked to learn a set novel, procedural computer-based tasks to mimic the scenario of an online learning course. Later they were brought back into the lab and asked to execute the same tasks learned days prior, but with varying review methods. From these measures, I gained a stronger understanding of what type of review style of previous work elicits the most relevant contextual memories. Click Here for PDF version. (video)