Picturing Better Materials: Normative Data on a Novel Illustration Set for Scientific and Clinical Use.

Abstract

Purpose: Picture stimuli are essential materials for language research and clinical practice. Here, we generated a modern, full-color set of 310 illustrations representing a carefully designed, culturally sensitive list of imageable nouns. We normed the images in a diverse sample of healthy adults, so the images can be used in various populations, including older adults.

Method: We recruited a diverse online sample of 200 adults ages 19-76 years. Participants typed a name for each picture and reported how familiar they were with the item (familiarity) and how well the illustration matched their mental image of the item (image agreement). We assessed relationships among these three measures, between these measures and word features (e.g., frequency, length), and between these measures and demographic characteristics of our sample.

Results: Two hundred ninety-seven of 310 items had 70% or higher name agreement among participants. Most items had good to excellent image agreement and familiarity. The image measures showed expected relationships with relevant word features (e.g., frequency, length). Older age was associated with higher image agreement and familiarity but not written naming accuracy. As a group, Black participants demonstrated lower written naming accuracy than White and mixed race participants. Education, sex, and self-reported multilingualism were not significantly related to image measures.

Conclusions: We generated a novel set of illustrations with strong name agreement, familiarity, and image agreement, which are suitable for a variety of uses in research and clinical settings. Our normative data suggest a future need for item-level analysis to explore variability in performance across different racial groups.

Supplemental material: https://doi.org/10.23641/asha.26321926.