ABSTRACT
This chapter concerns an experimental procedure that permits variation on all three dimensions, and with theoretical methods to interpret the reading time data acquired with that procedure. In the basic procedure, subjects press a button to display one word at a time in the center of a screen. When the computer displays each word, it extinguishes the preceding word and records its reading time (RT). The word-by-word paradigm is an “on-line” procedure for indexing perceptual encoding during reading. That is, the RT duration data are obtained while the subject is perceiving each individual word. With most other procedures, indirect inferences about perception are made based on subsequent memory and comprehension responses obtained at time delays after the sentence is read. The word-by-word RT procedure yields very reliable data. In comparison to other experimental procedures used to study reading, the variability is quite low. The button-pressing feature of the word-by-word procedure may be criticized as “unnatural”.
