When guessing is better than thinking: Multiple bases for frequency judgments

There is an ongoing debate whether frequency judgments are based on mental magnitudes reflecting prior on-line recording of frequencies or on recall content available at the time of judgment. We conducted four experiments to demonstrate that task characteristics can determine which kind of information (mental magnitude vs. recall content) is used in frequency judgment. The results of the experiments convergently show that frequency judgments are likely to be based on mental magnitudes under time constraints or when participants are instructed to give spontaneous judgments. In contrast, when participants are paid contingent upon their performance or are instructed to be accurate, frequency judgments reflect recall content.