Virtual Chinrest: Viewing Distance

Visual angle, the angle a stimulus subtends at the eye, depends on the stimulus's physical size and on the viewing distance, and neither is known in an online study. This task implements the virtual chinrest of Li, Joo, Yeatman and Reinecke (2020), which estimates both.

First, the participant matches an on-screen rectangle to a card in bank-card size, which gives the screen scale. Then the participant covers the right eye and looks at a black square with the left eye while a red dot moves to the left. The dot disappears when it enters the blind spot, the part of the visual field that falls on the optic disc, where the retina has no photoreceptors. The participant presses the space bar at that moment. Because this point lies at about the same angle from fixation in most people, the distance s between square and dot gives the viewing distance as d = s / tan(13.5°). The task averages five trials, shows the estimate, and lets the participant measure again if it seems wrong.

In Li et al.'s (2020) laboratory validation, the estimate of a 53 cm distance had a mean absolute error of 2.36 cm, and across distances of 43–66 cm and two screen sizes the mean absolute error was 3.25 cm. With a related blind-spot procedure, in which distances were read off a regression fitted to other participants' data rather than computed with a fixed 13.5°, Brascamp (2021) found a mean absolute error of 5.5 cm against participants' own tape-measure readings.

The result is available to later screens and to the tasks that follow it in the study flow: calc(2 * var(--deg)) makes a stimulus 2° wide. If Screen Calibration: Card Matching or Screen Calibration: Ruler Measurement ran earlier in the participant's session, this task uses that screen scale and skips its own card step, provided the page has not been reloaded and the device pixel ratio is unchanged. Added again later in the study flow, for example between two experiment tasks, it measures the distance again without repeating the card step.