Do coloured filters benefit reading beyond placebo?

  • P J Cowan

Student thesis: Doctoral ThesisDoctor of Philosophy (PhD)


Individually chosen coloured filters (overlays or lenses) are claimed to benefit reading for visual stress sufferers, but this could be attributed to a placebo effect. Previous research lacks both suitable measures of reading benefit and methods of controlling for placebo effects, and the effect of contrast on visual stress is relatively unexplored.

A ‘cloze’ task was developed to measure the reading benefit of filters. Additional experiments tested whether chosen coloured filters improved reading performance, visual search or single word reading of filter users; placebo effects were controlled for using change blindness and Stroop Interference. The effect of leading instructions on performance of the reading task used to diagnose visual stress (WRRT) was investigated, and a visual stress questionnaire was administered. The effects of contrast on preference ratings for people who reported visual stress were measured. Lastly, the effect of reducing text contrast on filter users’ reading performance was quantified.

The cloze task demonstrated no reading benefit of a chosen coloured filter. Leading instructions were found to improve WRRT reading speeds with a chosen coloured filter but not the cloze task. Participants who reported visual stress had improved cloze reading performance with any filter and reported that they found text easier to read on blue or grey paper rather than white. However, filter users did not improve their reading speed with a reduced text contrast setting.

This thesis presents no new evidence to support the view that chosen coloured filters are needed to reduce visual stress beyond placebo. However, effects of placebo were observed with a chosen filter on the WRRT, casting doubt on this method for the diagnosis of visual stress. Some data reported here indicate that reduced contrast may reduce visual stress.
Date of Award28 Nov 2019
Original languageEnglish
Awarding Institution
  • University of Bristol
SupervisorNicholas E Scott-Samuel (Supervisor) & Clive Frankish (Supervisor)

Cite this