Dispatch Console Binding Load
You are a human-factors researcher evaluating a new dispatch console for a parcel airline. The console shows eight status icons in fixed positions. In the "feature" condition, icon color changes every 2.5 seconds and dispatchers report which colors changed. In the "binding" condition, the same colors change but dispatchers must report which color appeared at which location. Accuracy is 92% with three changing items, 71% with five, and 54% with seven. Most errors in the binding condition are swaps: the correct colors are reported but assigned to the wrong icons. Adding a verbal shadowing task leaves the feature condition unchanged but drops binding accuracy from 71% to 48%. Adding spatial tapping also impairs binding, though less than shadowing. You must advise the design team on how many simultaneous changes the console should allow and how to present them.
1. What do the capacity curve and the swap errors indicate about visual working memory in this task?
Read the worked answer
The steep drop from 92% at three items to 54% at seven is consistent with a capacity limit around three to four object representations, not unlimited storage. The swap errors are the key: colors are retained but bound to the wrong locations, which indicates a binding failure rather than loss of color features. That pattern fits object-based visual working memory, where features must be integrated into location-specific episodes. The verbal shadowing result suggests participants were partly using verbal labels to support binding, or that the task recruits domain-general attention; either way, binding is the fragile stage. Spatial tapping's effect supports a visuospatial component. The practical reading is that the console is overloading binding, not color discrimination.
2. The design team proposes allowing six simultaneous changes to save screen space. What is your recommendation, and what presentation changes would reduce binding load without adding screens?
Read the worked answer
Recommend no more than three simultaneous changes per 2.5-second cycle, and preferably stagger changes so they do not all occur in one frame. Six changes exceeds the observed capacity and will produce swap errors that look like operator mistakes. Reduce binding load by keeping status location constant and changing only one feature at a time; grouping related icons spatially so a single spatial chunk carries several statuses; using a transient highlight plus a persistent text tag so the operator does not have to hold the binding; and avoiding color-only coding for critical states by adding shape or position. If six changes are unavoidable, slow the cycle or add an auditory alert that cues the location. Evaluate any change with the same binding task, not just subjective preference.
3. How would you distinguish a visual working-memory limit from a long-term visual memory problem in this paradigm, and why does it matter for training?
Read the worked answer
Visual working memory is tested within a trial with a brief delay and no opportunity for rehearsal or consolidation; long-term visual memory is tested after minutes or days and benefits from repetition, meaning, and sleep. Here, the 2.5-second update and immediate report keep the load in working memory. To test long-term memory, show a static console layout, remove it, and test recognition after 20 minutes; if that is intact while on-line binding fails, the problem is capacity, not storage. It matters because training cannot expand a capacity limit; it can only build chunking, consistent spatial layouts, and external cues. If the problem were long-term memory, rehearsal and documentation would be the target. The distinction prevents the team from buying memory training when the fix is interface design.