Abstract
Since development of computer technology has more frequently caused people to view the data graphically displayed on a computer display and to make various judgments, understanding of how human judgments are affected by display formats has become much more important. In this study, two experiments were performed in order to investigate how human judgments of relatedness of variables presented on scatterplots were affected by the number of data and the presentation of regression line. The major results were (a) in the case that the two variables possessed relatively high correlation, human judgment approached the value of correlation coefficient when viewing a large number of data than when viewing a small number of data; (b) presentation of a regression line on the scatterplot had no significant effect on subjects' judgments of relatedness of two variables.
