Abstract

Letter optotypes such as Sloan letters (NAS-NRC, 1980) are typically used in the clinical assessment of spatial vision, and scoring procedures are usually based on the number of individual letters that are identified correctly (e.g., Bailey et al., 1991; Elliott et al., 1991). This approach assumes that all Sloan letters have approximately equal identifiability, but there is some disagreement on this point (e.g., Elliott et al., 1990; Robson et al., 1990; Ferris et al., 1993). This may be due to the method used to assess identifiability. Typically, previous studies have assessed letter identifiability by measuring percent correct at threshold; relatively few have examined thresholds per se. Depending on the nature of the underlying psychometric function, these two approaches may not provide equivalent estimates of relative letter identifiability.

© 1996 Optical Society of America