Eliciting Probabilities in the Field : Beads versus Sliders

Abstract

Elicited probabilities have been shown to be effectively in explaining and predicting behavior. Their elicitation is typically facilitated by visual aids, such as beads or beans. This study, introduces a novel aid, consisting of a cardboard-based slider, to facilitate probability elicitation among populations with low levels of education. Unlike discrete aids, such as beads, the slider allows respondents to report probabilities on a continuous scale, which is particularly advantageous for statistical inference and econometric modeling. To compare the effectiveness of these aids, the study conducted four survey experiments: two in Nigeria, one in Honduras, and one in Spain. These experiments assessed the validity of elicited expectations using both visual aids. The study found no significant differences in participants’ understanding of basic probability concepts or in the time required to complete the task. However, it found that the slider yielded more precise measures (lower variance) than the beads. These results suggest that comprehension remains consistent across both aids, but that the slider improved the precision of subjective probability elicitation in the field.