The Distinction between the Rod-and-Frame Illusion and the Rod-and-Frame Test

Publisher Website
Google Scholar
Abstract
Errors in vertical settings of a test rod occur when the rod is enclosed in a laterally-tilted square-outline frame. The majority of previous experiments which have investigated this rod-and-frame effect have used a single frame tilt, usually 28°, and have tabulated errors as average unsigned deviations from gravitational vertical. Evidence is presented that, when the illusion is measured by taking algebraic differences between constant (signed) errors made with and without the frame being present, illusions occur in the direction of frame tilt for frame tilts up to about 25° from vertical (repulsion effects) but that directionally opposite illusions (attraction effects) occur for frame tilts between 25° and 45°. At the frame tilts used most frequently in previous studies (25° to 30°) little or no illusion occurs. A distinction is drawn between the rod-and-frame illusion (RFI), which has an angular function similar to the simple tilt illusion and aftereffect, and the rod-and-frame test (RFT), which uses unsigned deviations from vertical as its measure of error and which probably bears little or no relationship to the RFI.