A ruler-based technique to rigorously control the size of visual stimuli for online psychological experiments

This study proposes a rigorous method to control the size of visual stimuli for online psychological experiments. Our method employs ruler-based calibration to estimate the on-screen millimeter-to-pixel ratio by applying the least squares method to repeated measurements made by participants of randomly generated line segment lengths. Calibration success can be objectively verified by assessing both measurement errors during the test phase and the residual variance of a linear regression model. Our simulation results demonstrate that an upper bound of the standard deviation of estimation errors was 0.007 mm/px under the proposed calibration criteria. Demonstrating the method’s feasibility, implementation results showed that the majority of participants (N=100) were able to complete calibrations within a few minutes. Our method may facilitate rigorous studies of psychological phenomena that are sensitive to visual-stimulus size, such as visual perception and cognition, conducted in online environments. A demo and a JSON file for implementing the method with lab.js are available at https://doi.org/10.17605/osf.io/zeyxp.