Screening Designs Used to Estimate the Relative Effects of Display Factors on Dial Reading
通过两个实验测试筛选设计能否可靠估计多个自变量对指针式仪表读数速度和准确性的相对影响，发现数字排列影响最大，插值和刻度单位长度影响较强，其他因素影响弱或可忽略。
Two experiments were conducted to determine if screening designs could be used to obtain reliable estimates of the relative strength of effect of several independent variables. Specifically, the effects of eight independent variables on the speed and accuracy of reading moving-pointer, fixed-scale dials were tested. Additional objectives were to determine if the black-white relationship of the scale markers and dial background affects dial-reading performance or subject preference. The percentage of the variance accounted for by each main effect was very similar across screening designs. The progression of the numbers on the scale had by far the strongest effect. Interpolation and scale unit length also had relatively strong effects, while scale orientation, marker width, pointer design, clutter, and scale number location had weak to trivial effects. Black-on-white and white-on-black dials were read at about the same speed, but more errors were made while subjects read the black-on-white, even though they preferred the black-on-white dials.