Time-Accuracy Elimination

Problem
Some evaluations come down to time and error performance. However, it is often difficult to balance both measurements given the individual differences of participants (the so-called "time-accuracy tradeoff"). One participant may be very thorough and score few errors at the cost of high completion times, whereas another participant may quickly solve tasks while incurring many errors. This remains an issue even with very specific instructions.

Solution
Design experimental tasks so that one of time or error measurements are eliminated. An example of eliminating error would be for a visual search task, where you can design the task so that the participant is not allowed to answer with an incorrect target; the trial only ends when the correct target is selected. Analogously, to eliminate timing, either give a specific time limit (say, 10 seconds), or give no time limit at all to find the target, but allow only one click.

Consequences
Eliminating time or error lets you analyze only one metric, and will give more definite answers on the impact of the conditions. On the other hand, it may cause you to miss nuanced tradeoffs between time and error.

Examples
In a recent study on occlusion management for tabletop interfaces (Javed et al. 2011), Javed et al. designed a task where participants were asked to recall a sequence of images so that out-of-order selections were not possible. In other words, errors were eliminated and only timing was analyzed.