What Causes Garner Interference? Toward a Sequential Feature Binding Account of the Effect

Four experiments were performed with the aim of testing the sequential episodic binding account. Experiment 1 and Experiment 2 used Munsell dimensions of chroma and value. Experiment 3 and Experiment 4 harnessed width and height of rectangles.

The results of all four experiments strongly supported the predictions of the sequential binding account: (a) Garner interference significantly decreased as a function of inter stimulus interval (ISI), a variable that has been shown to reduce binding strength, (b) binding effects (partial repetition costs) were documented in filtering, and they decreased as a function of ISI, (c) the size of Garner interference was strongly correlated with the size of binding effect (partial repetition costs) in filtering, and (d) feature binding exceeded a critical quantity predicted by the theory to be violated with integral (but not separable) dimensions.