Exploring a Counterintuitive Finding with Methodological Implications

In the present experiment, 107 undergraduates judged the size of 9 or of 221, using either a numerical scale from 1 to 10 or a continuous line, both of which were anchored by the labels low and high. Replicating previous research with the numerical scale and more extreme labels (very very small, very very large), it was found that 9 was rated greater than 221. In addition, from reports of the meaning of the end points of the scales, it was concluded that 9 invoked a smaller context than 221, at least for some participants. These results add to concerns that between-subject designs are not free of context effects that may lead to anomalous outcomes.