Common factors versus uncommon variability in comparisons of the Poggendorff and Müller-Lyer illusions

Abstract

Evidence for a common contributing factor in the Miiller-Lyer and PoggendorfT illusions is reviewed and complemented by correlational studies of the two illusions. However, these findings are compromised by consistent differences in the variability of the two illusory effects. In general, the variability associated with judgments of linear extent is consistently greater than the variability associated with alignment judgments. These findings were obtained with different figures, different stimulus parameters, and different response procedures. The difference in variability suggests problems for the notion of a common causal factor in illusions of linear extent and of direction.

The long and short of it: physical anchoring effects

“…In some cases, participants are asked to perform a single adjustment procedure (e.g., all participants shorten lines; Predebon, 1996;Redding & Hawley, 1993;Warren & Bashford, 1977), whereas in other cases some participants increase a low stimulus, others decrease a high stimulus, and an average of these conditions is used for analysis (Day & Dickinson, 1976;Luo & Wang, 1997). This latter procedure suggests at least a tacit awareness that anchors and starting positions might bias judgment, but although there have been several methodological investigations of magnitude re-creation (e.g., Schiffman & Griest-Bousquet, 1993;Stevens & Guirao, 1963), such prior research has typically ignored the potentially striking and systematic effects of starting points.…”