Do Pattern-Focused Visuals Improve Skin Self-Examination Performance? Explicating the Visual Skill Acquisition Model

Abstract

Skin self-examination (SSE) consists of routinely checking the body for atypical moles that might be cancerous. Identifying atypical moles is a visual task; thus, SSE training materials utilize pattern-focused visuals to cultivate this skill. Despite widespread use, researchers have yet to explicate how pattern-focused visuals cultivate visual skill. Using eye tracking to capture the visual scanpaths of a sample of laypersons (N = 92), the current study employed a 2 (pattern: ABCDE vs. ugly duckling sign [UDS]) × 2 (presentation: photorealistic images vs. illustrations) factorial design to assess whether and how pattern-focused visuals can increase layperson accuracy in identifying atypical moles. Overall, illustrations resulted in greater sensitivity, while photos resulted in greater specificity. The UDS × photorealistic condition showed greatest specificity. For those in the photo condition with high self-efficacy, UDS increased specificity directly. For those in the photo condition with self-efficacy levels at the mean or lower, there was a conditional indirect effect such that these individuals spent a larger amount of their viewing time observing the atypical moles, and time on target was positively related to specificity. Illustrations provided significant gains in specificity for those with low-to-moderate self-efficacy by increasing total fixation time on the atypical moles. Findings suggest that maximizing visual processing efficiency could enhance existing SSE training techniques.

Funding

Research reported in this publication was supported by the National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health under award number 1DP2EB022360-01 (PI: J. Jensen).

Notes

¹ Clinically, nevus or lesion would be preferred terms. However, moles will be used throughout the article.

² Statistical power calculations were executed using G*Power 3.1.5 (Faul, Erdfelder, Buchner, & Lang, 2009). The analytical approach utilized two-way ANOVAs and moderated mediation analyses. For ANOVA analyses, effect size standards are small (f² = .10), medium (f² = .25), and large (f² = .40). For the moderated mediation model, effect size standards are small (f² = .02), medium (f² = .15), and large (f² = .35) (Faul et al., 2009; Faul, Erdfelder, Lang, & Buchner, 2007). Achieved power for the two-way ANOVAs was excellent for the detection of large effects (.97), good for the detection of medium effects (.66), and poor for the detection of small effects (.16). Achieved power for the moderated mediation models (with six predictors in the model) was excellent for the detection of large effects (.99), good for the detection of medium effects (.81), and poor for the detection of small effects (.14).

Additional information

Funding

Research reported in this publication was supported by the National Institute of Biomedical Imaging and Bioengineering of the National Institutes of Health under award number 1DP2EB022360-01 (PI: J. Jensen).