Fixation location effects on fixation durations during reading: An inverted optimal viewing position effect

Françoise Vitu, George W. McConkie, Paul Kerr, J. Kevin O'Regan

Research output: Contribution to journalArticlepeer-review


Previous research has found that words are identified most quickly when the eyes are near their center (the Optimal Viewing Position effect). A study was conducted to determine whether this same phenomenon is observed during reading, as revealed by a relationship between fixation position in a word and the duration of the fixation. An analysis of three large existing corpora of eye movement data, two from adults and one from children, showed a surprising inverted Optimal Viewing Position curve: mean fixation duration is greatest, rather than lowest, when the eyes were at the centers of words. From this phenomenon, we suggest an alternative explanation to the fixation duration trade-off effect in word refixations [O'Regan & Lévy-Schoen, Attention and performance XII: the psychology of reading (1987)]; the phenomenon also contradicts expectations of both oculomotor and cognitive theories of eye movement control. Attempts to test alternative explanations led to the discovery of another phenomenon, the Saccade Distance effect: mean fixation durations vary with the distance of the prior fixation from the currently-fixated word, being longer with greater distances. The durations of fixations in reading are complexly determined, with influences both from language and perceptual/oculomotor levels.

Original languageEnglish (US)
Pages (from-to)3513-3533
Number of pages21
JournalVision Research
Issue number25-26
StatePublished - Dec 10 2001


  • Eye movements
  • Fixation duration
  • Fixation location
  • Reading
  • Strategy
  • Word identification

ASJC Scopus subject areas

  • Ophthalmology
  • Sensory Systems


Dive into the research topics of 'Fixation location effects on fixation durations during reading: An inverted optimal viewing position effect'. Together they form a unique fingerprint.

Cite this