Using an experimental mixed design, this study compared the traditional paper-and-pencil method for evaluating teaching with the online method. Replicating previous findings, the comparison revealed similar evaluation means of the two methods. However, the stability of teaching evaluations using paper-and-pencil twice was substantially higher than the corresponding stability using different methods--online and paper-and-pencil. One possible explanation for this finding is the different visual presentation of the scales: a typical form of the paper-and-pencil method presents the scale horizontally, enabling the subjects to examine the profile of their answers that might result in an artificially lower variability of the evaluations. In contrast, an electronic answering form can abolish this artificial answering effect. (Contains 2 endnotes.)