An explicit no response instead of Time-out in automated visual-field testing

• Published in: Graefe's archive for clinical and experimental ophthalmology 2001-03, Vol.239 (3), p.173-181
• Main Authors: LUTZ, Susanne, DIETRICH, Traugott J, BENDA, Norbert, SELIG, Bettina, STRASBURGER, Hans, SCHIEFER, Ulrich
• Format: Article
• Language: English
• Subjects: Adult; Biological and medical sciences; False Positive Reactions; Female; Humans; Investigative techniques of ocular function and vision; Investigative techniques, diagnostic techniques (general aspects); Male; Medical sciences; Ophthalmology; Predictive Value of Tests; Psychometrics - methods; Reproducibility of Results; Sensitivity and Specificity; Surveys and Questionnaires; Visual Field Tests - methods; Visual Fields - physiology
• ISSN: 0721-832X; 1435-702X
• DOI: 10.1007/s004170000243

To evaluate the effect of response-acquisition technique on psychometric performance in visual-field testing, the conventional one-button yes/time-out method was compared with a two-button yes/no method for responding whether or not the stimulus was detected. There are a number of situations in which the single-button technique leads to ambiguous results. In this study, we thus expected the yes/no method to reduce tendencies towards habituation and automatic responding. Our hypothesis was that the two-button technique could reduce the rate of erroneous responses. Luminance-difference sensitivity for bright stimuli (32') on a photopic background was evaluated at 26 locations within the central visual field (30 degrees) using a specially equalised video display unit and a modified 4/2-dB staircase strategy (six reversals, maximum-likelihood threshold estimation). Sixty-one ophthalmologically normal subjects (aged 20-30 years) were examined twice with each method. Mean sensitivities with the two-button yes/no method were found to be, on average, 0.13 dB above those measured with the one-button yes/time-out technique--a difference without clinical relevance. Within-subject variability did not differ between the two methods. However, the less intuitive two-button yes/no method had a slightly higher number of false responses in catch trials. Compared to the conventional one-button yes/time-out method, the two-button yes/no method in normal young subjects thus showed little difference in mean sensitivities and equivalent within-subject variabilities. Concerning our initial hypothesis, the yes/no method is of somewhat higher complexity and is not able to reduce the rate of erroneous responses. The one-button yes/time-out method fared a little better in error rate. In summary, the yes/no method is an alternative and additional possibility of response acquisition in visual-field testing, which is worthy of being tested in a clinical study with elderly subjects.