Visual Photometry: Testing Hypotheses Concerning Bias and Precision

Volume 51 number 1 (2023)

Download this article (pdf)

Alan B. Whiting

Department of Physics and Astronomy, University of Birmingham, Edgbaston, Birmingham BT2 5TT, UK; abw@star.sr.bham.ac.uk

Abstract

Visual photometry, the estimation of stellar brightness by eye, continues to provide valuable data even in this highly-instrumented era. However, the eye-brain system functions differently from electronic sensors and its products can be expected to have different characteristics. Here I characterize some aspects of the visual data set by examining ten well-observed variable stars from the AAVSO International Database. The standard deviation around a best-fit curve ranges from 0.14 to 0.34 magnitude, smaller than most previous estimates. The difference in scatter between stars is significant, but does not correlate with such things as range or quickness of variation, or even with color. Naked-eye variables, which would be expected to be more difficult to observe accurately, in fact show the smallest scatter. The difference between observers (bias) is less important than each observer’s internal precision. A given observer’s precision is not set but varies from star to star for unknown reasons. I note some results relevant to other citizen science projects.