Adapting a Psychophysical Method to Measure Performance and Preference Tradeoffs in Human-Computer Interaction