Evaluating text, icon, and graphic nutrition labels: An eye tracking experiment with Latino adults in the US

Abstract

The US Food and Drug Administration (FDA) is developing front-of-package nutrition labels for packaged foods. Identifying the most promising type of label among Latino adults could inform federal regulation, given high rates of diet-related disease in Latino populations. Additionally, exploring English-language label effects among populations with limited English proficiency could inform equitable label design. We examined whether text, icon, or graphic nutrition labels attract attention among Latino populations and whether label effects differed by English proficiency. In 2023, we recruited 63 adults in North Carolina identifying as Latino/a/é (hereinafter “Latino”); 48% had limited English proficiency. Participants viewed four labels on a can of soup in random order: a barcode label (control) and text, icon, and graphic labels reading, “WARNING: High in sodium.” Eye trackers measured time spent viewing the label (dwell time), number of times viewing the label (fixation count), and time to first fixation on the label. A survey assessed secondary outcomes. Dwell time was highest for the graphic label (mean=2.58 seconds (s)), followed by icon (mean=2.34s), text (mean=1.94s), and control labels (mean=.96s; p for each label vs. control <.001). The impact of label type on dwell time did not differ by English proficiency (p=.669). Fixation count was highest for the graphic label, followed by the icon, text, and control labels (p for each label vs. control <.001). Participants viewed the graphic and text labels more quickly than control (ps=.01). Self-reported attention, perceived message effectiveness, and understandability were higher for text, icon, and graphic labels than control (all p<.001 vs. control). This study suggests that front-of-package labels signaling that foods are high in nutrients of concern can attract consumers’ attention, especially when the labels include images or icons.