The straight truth: measuring observer attention to the crooked nose

Abstract

Objectives/hypothesis: Quantify attentional distraction to crooked noses pre- and postoperatively as compared with normal noses by using an established metric of attention in a pilot study.

Study design: Prospective, randomized, controlled experiment with crossover.

Methods: An eye-tracker system was used to record the eye-movement patterns, called scanpaths, of 40 naive observers gazing at pictures of faces with crooked noses preoperatively or postoperatively and pictures of faces without a crooked nose included as "normals." The fixation durations within the nasal area for each group of faces presented were compared.

Results: A mixed-design univariate analysis of variance was performed to test the hypothesis that mean fixation times in the nasal region varied by face group. The results were highly statistically significant, F(2,116) = 20.28, P = .000, η(2) = 0.029. Marginal means were calculated for each nasal area of interest group with confidence intervals (normal, 2.32 [2.26-2.38]; preoperative, 2.66 [2.58-2.75]; postoperative, 2.43 [2.35-2.51]). Post hoc testing with Bonferroni correction for three comparisons showed differences between the normal and preoperative groups (χ(2) 41.38, P = .000) and between the preoperative and postoperative groups (χ(2) 14.41, P = .000) but not between the normal and postoperative groups (χ(2) 4.19, P = .12).

Conclusions: There were highly statistically significant differences in attention paid to the nasal area of crooked noses preoperatively and postoperatively, and there were no differences in attention to the nasal area between the postoperative noses and the normal noses. This represents a novel method for objectively evaluating attention and success of surgical procedures to minimize the appearance of deformities.