The "-2>2" Effect In Size Judgment

Birnbaum(1999)found an intriguing phenomenon: He asked two groups of subjects to judge how large “9” was and how large “221” was respectively on the scale of the same style.The judged size of “9 “was larger than that of “221”! This is the so-called “9 > 221” effect.Furthermore,Leong(2019)ingeniously found “9 > 009” effect and “-2 > 2” effect.Nevertheless,there are at least one of the following problems in their experiments.(1)Their subjects were factory workers.Their education levels were relatively low,and they were possibly not so careful in dealing with questions about numbers.(2)Their subjects are westerners.It is well known that Chinese are much better at numeracy than westerners.Therefore,it is a big question whether Chinese responses to these numeracy questions in the same way as the westerners.(3)One scale used in Leong(2019)was strange.There was still a long line segment at the left end of this scale,and results from this strange scale may be hard to be generalized to the normal scales in normal research.To overcome these problems,we recruited people with the high educational level and high numeracy skills(Chinese university students)as subjects,and used normal scales to do a series of experiments to test whether the above intriguing phenomena can reappear in rigorous experiments.Taken together,in a series of strict experiments(using subjects with the high educational level and high numeracy skills,and using normal scales),we replicated all the three phenomena:“9 > 221” effect,“9 > 009” effect,and “-2 > 2” effect.These results re-remind researchers who use scales as tools to avoid subtle but vital compounded variables in their research to ensure the reliability of their research conclusions.