Abstract
Blinding is an essential part of many randomized controlled trials. However, its quality is usually not checked, and when it is, common measures are the James index and/or the Bang index. In the present paper we discuss these two indices, providing examples demonstrating their considerable weaknesses and limitations, and propose an alternative method for measuring blinding. We argue that this new approach has a number of advantages. We also provide an R-package for computing our blinding index.