Abstract
Studies on recreationists' norms from different sites have reported different percentages of recreationists who have norms, leading some to question whether such differences result from different question formats. A random sample of 365 boaters on the Clackamas River in Oregon was used to explore the effects of two formats on the presence of norms for three social impacts (encounters with other boaters, time waiting at launch site, and percentage of time in sight of other boaters) and on the level of agreement among those boaters who had personal norms. In a two-choice format, boaters could state a norm or indicate that the impact did not matter to them. A three-choice format added a third option: "The impact matters to me, but I cannot give a number.'' Between 16 and 39% of boaters selected the third option, depending on the impact in question. The three-choice format resulted in fewer boaters saying they had norms, but also lowered the variability among those who gave norms.