Entry and selection methods for specifying dates

Experienced and inexperienced computer users specified event dates using seven different human-computer interaction methods (three entry and four selection methods). Both groups were faster and more accurate when using the three entry methods than the four selection methods. The rank order of performance with these methods was about the same for both groups of participants. The number of keystrokes required by each method was a good predictor of participants' performance time. It was found that decomposing selection tasks into separate fields is advisable.