PDF Archive

Easily share your PDF documents with your contacts, on the Web and Social Networks.

Share a file Manage my documents Convert Recover PDF Search Help Contact

Reliability and Validity of Eye Movement Measures of Reading.pdf

Preview of PDF document reliability-and-validity-of-eye-movement-measures-of-reading.pdf

Page 1 2 3 4 5 6 7 8 9 10 11

Text preview



from .62 to .87. Similarly, Frandsen 6 obtained reliabilitycoefficients ranging from .78 to .88 for adult reading of prose,
and from .55 to .90 for reading objective questions. All the
above coefficients were obtained by the odd-even or split-half
method. Test-retest reliabilities are needed.
There are several aspects to be considered in studying the
validity of eye-movement measures of reading. These have
been summed up by Tinker.8 An evaluation of the measures
revealed that perception time and fixation frequency are good
measures, regression frequency a fair, and pause duration a
poor measure of reading speed. It was further shown that
the direct method of observing eye movements is quite inaccurate and only useful to reveal general trends.
Validity of eye-movement measures has been computed by
correlating the measures with achievement on standardized
reading tests. By this method Litterer 7 obtained for adults
coefficients ranging from — .27 to — .51. (The negative
coefficients indicate that the reading score increased as the
fixation frequency decreased, etc.) Comparable results from
Eurich 8 ranged from — .02 to — .24 for adults, and from
— .11 to — .55 for children. It is clear from these coefficients
that validity with achievement test scores as criteria is low.
As will be shown later, this is an unsatisfactory technique of
determining validity. Other methods of attacks should be
The purpose of the present study is twofold: (1) To determine by the test-retest method the reliability of eye movements in reading materials of various lengths. (2) To investigate the validity of eye-movement measures by studying
adaptation to the artificial laboratory situation, and by
computing validity coefficients (a) with scores on standardized
reading tests as criteria and (b) with performance on strictly
comparable reading material as criterion.
•Frandsen, A., An eye-movement study of objective examination questions.
Getul. Psycho!. Monog., 1934, x6, No. 2, 79-138.
•Tinker, M. A., Eye movement, perception and legibility in reading. Psychol.
Bull., 1936, 33, 275-490.
Litterer, O. F., op. dt.
* Eurich, A. C , op. cit., p. 118 and p. 380.