Illumination frame of reference in the object-reviewing paradigm: A case of luminance and lightness

Abstract

The present study combines the object-reviewing paradigm (Kahneman, Treisman, & Gibbs, 1992) with the checkershadow illusion (Adelson, 1995) to contrast the effects of objects' luminance versus lightness on the object-specific preview benefit. To this end, we manipulated objects' luminance and the amount of illumination given by an informative background scene in experiments. In line with previous studies (Moore, Stephens, & Hein, 2010), there was no object-specific preview benefit when objects were presented on a uniformly colored background and luminance switched between objects. In contrast, when objects were presented on the checkershadow illusion background which provided an explanation for the luminance switch, a reliable object-specific preview benefit was observed. This suggests that object correspondence as measured by the object-reviewing paradigm can be influenced by scene-induced, perceived lightness of objects' surfaces. We replicated this finding and moreover showed that the scene context only influences the object-specific preview benefit if the objects are perceived as part of the background scene.

Figure 1

The checkershadow illusion first published by Adelson in 1995. Left: The checks A and B are equiluminant but their perceived lightness differs. Right: Superimposition of two bars in the same shade of gray destroys the illusion. ©1995, E.H. Adelson

Figure 2

Mean reaction time (RT in ms) error rates (Error in %) as a function of Congruency (same object vs. different objects) and Background (plain gray vs. checkerboard) in Experiment 1. Error bars show the mean ± standard error using the algorithm for within-subject designs recommended by Cousineau (2005).

Figure 3

Mean reaction time (RT in ms) error rates (Error in %) as a function of Congruency and Luminance in Experiment 2. Error bars show the mean ± standard error using the algorithm for within-subject designs recommended by Cousineau (2005).

Figure 4

Mean reaction time (RT in ms) error rates (Error in %) as a function of Congruency (same object vs. different objects) and Luminance (no switch vs. switch) in Experiment 3. Error bars show the mean ± standard error using the algorithm for within-subject designs recommended by Cousineau (2005).

Figure 5

Mean reaction time (RT in ms) and error rates (Error in %) as a function of Congruency (same object vs. different objects) and Luminance (no switch vs. switch) in Experiment 4. Error bars show the mean ± standard error using the algorithm for within-subject designs recommended by Cousineau (2005).