The effectiveness of two five-point Likert response scales with different verbal anchors was compared. Four anonymous written questionnaires addressing characteristics of effective schools were developed. Each questionnaire was prepared in two formats, one with each response scale. The 4 questionnaires were composed of between 140 and 200 items. A balance of positive and negative items was used. Each of 539 public school teachers completed both formats, with the order of administration randomized. The two response scales were found to be similar in terms of average required completion times, reliabilities, and respondents' written comments, but they differ in non-response rates. These results suggest a potential series of investigations that instrument developers may wish to use in developing questionnaires. It appears feasible and worthwhile to include comparison of alternate response scales in the pilot survey. Seven graphs present study data, and there is a 17-item list of references. (SLD)