Abstract
Laboratory studies of warning symbol signs have been shown to underestimate legibility distances by up to a factor of two when compared with field studies. However, the research reported here suggests it may be more than just experimental settings contributing to disparity in research findings. Using a group of old and young drivers, six warning symbol signs were investigated in both a field and laboratory setting. The six signs consisted of two each in three categories of spatial frequency content. In both settings, legibility distances, the distance at which the sign is correctly identified from a menu, were collected with six trials per sign per subject per condition. As a major part of this research, a new laboratory siimulation technique was developed, a methodology which optimized factors criticized in earliler studies, thus increasing fidelity. Correlation coefficients between laboratory and field legibility distances were very promising. The newly developed laboratory dimulation was a successful first step at correcting problems associated with laboratory studies in the past. Large within-subject variablility, evident in both age groups, was found to overshadow the contributions of experimental variables. No studies have reported the finding of such large within-subject variance in similar research. Identification of this substantial within-subject variability led to an anyalysis of the variation in distance estimation if different strategies for trials are used. Using only the first, or the minimum, or the first two give very different answers. It is argued that recommended distances at which signs are placed should be determined from a "worst-case" scenario of the minimum distance with tolerances constructed from the other trials. This premise requires a reexamination of research methodologies for determining guidelines for sign placement, the primary one is the necessity for more trials to account for within-subject variability.
Greene, Frances Anne (1994). A study of field and laboratory legibility distances for warning symbol signs. Texas A&M University. Texas A&M University. Libraries. Available electronically from
https : / /hdl .handle .net /1969 .1 /DISSERTATIONS -1554382.