@article{ref1,
title="Recognition time for letters and nonletters: effects of serial position, array size, and processing order",
journal="Journal of experimental psychology: human perception and performance",
year="1982",
author="Mason, M.",
volume="8",
number="5",
pages="724-738",
abstract="Three experiments address the question of whether end-effects (the advantage for the first and last letters in multi-letter arrays) are due to sensory or cognitive factors. The basic strategy used was to compare letters with symbol stimuli that do not lend themselves to top-down processing. Serial position functions using central fixation were obtained in the first experiment for letters, digits, and symbols. End-effects were present with both letters and digits, indicating that processes limited to word recognition cannot be responsible for end-effects. Five-symbol arrays, however, yielded U-shaped reaction time functions, with poorest performance at the initial and terminal positions. The effects of array size and retinal placement were investigated with letters and symbols in a second experiment. Multi-letter and symbol arrays differed primarily in that letters showed end-effects that were independent of retinal placement, whereas symbols did not. The conclusion that multi-letter and symbol arrays are not processed in the same way was tested in a third experiment. Using a paradigm that eliminated processing order as a variable, the experiment obtained identical effects of ordinal position within an array for letters, digits, and symbols. It is concluded that sensory factors cannot provide a viable account of letter end-effects and that letter and nonletter comparisons can improve our understanding of the sensory and cognitive factors involved in letter perception.<p /><p>Language: en</p>",
language="en",
issn="0096-1523",
doi="",
url="http://dx.doi.org/"
}