JRRD at a Glance
Models of reading performance in older adults with normal age-related vision
Rolf W. Nygaard, PhD, et al.
Reading rate response to changes in print size (data) was simulated using a sigmoidally shaped model of positive growth with which several reading rate characteristics could be calculated. When the length of the data set to be regressed was shortened, thereby increasing efficiency of data collection, our model of positive growth showed reduced per datum error. This finding suggests that, at least for some observers, reading performance is reduced at large print size. Shortening the data set length, however, alters the parameters of fit. A compound model of positive and negative growth showed improved fit in some observers, giving additional reading rate characteristics and a means of determining reduced reading performance at large print size. We considered the role that experimenter uncertainty might play in estimating characteristics like maximum reading rate and critical print size. The results highlight the objectivity and logic of regression when applied to the critical human activity of reading performance.
Volume 45 Number 6, 2008
Pages 901 — 910
View HTML ¦ View PDF ¦ Contents Vol. 45, No. 6