Background: Acuity for letter recognition is known to be worse when multiple letters are presented with narrow interletter spacings than with wide spacings. How would interletter spacing affect the kind of errors made by human subjects?
Methods: Five-letter strings that were randomly drawn from the 26 uppercase letters of the English alphabet were presented foveally to the subjects. The interletter spacings were 1.0 and 0.1 letter height. Letter confusion matrices were constructed from the data collected using these spacing conditions.
Results: Narrow- and wide-spacing letter strings produced different letter confusion matrices. Aside from the letter confusions that were shared by both wide- and narrow-spacing strings, narrow-spacing strings produced more random confusions and a set of unique letter confusions, which was not observed under the wide-spacing condition.
Conclusion: Increased random guessing and lateral interactions between features of neighboring letters can account for most of the acuity deterioration observed under the narrow-spacing condition.
Download full-text PDF |
Source |
---|---|
http://dx.doi.org/10.1097/00006324-200101010-00014 | DOI Listing |
Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!