Multiple reasons for ``inconsistent" preferences over risks have been proposed. A leading explanation, incompleteness, is notoriously elusive to measure with standard choice data. Therefore, we leverage two experiments to test the degree of indecisiveness in individuals' choices over lotteries. To elicit the degree of indecisiveness, we present subjects with a series of choice lists and allow them to defer some of their choices or explicitly randomize between binary alternatives. We also measure preferences under different frames, gains and losses. These procedures help elicit the range of choices for which subjects may be unsure. Eliciting ranges allows us to determine utility as a range rather than a single value. Further, we test whether subjects' preferences may be summarized by a multi-utility representation and exhibit indecisiveness. Our results suggest individuals are sensitive to framing and averse to reporting indecisiveness directly.