Matching related stimuli is a foundational skill upon which a wide variety of critical repertoires are constructed. For learners with complex communication needs, including many children with autism, learned relations between pictorial stimuli and their referents are the basis for communication (Andzik et al., 2017; Crowe et al., 2021). Augmentative and alternative communication (AAC) tools enable such learners to engage in verbal behavior through the use of two-dimensional images, hereafter, pictures. These pictures are used to indicate real-world objects, persons, or activities to a communication partner. The form of these pictures varies widely; some pictures share many stimulus features with their real-world referents (e.g., photographs), whereas others share no discernable features (e.g., alphanumeric characters; McIlvane et al., 1993). Some children struggle to learn relations between pictures and their real-world referents, which can impede their access to such communication tools (Lionello-DeNolf & McIlvane, 2016).The default practice in stimulus selection for use with AAC devices at present is to select generic computerized line drawings from standard image libraries (such as those included with the AAC software). The use of full color photos of stimuli in the learner's natural environment is generally reserved for learner-specific stimuli, including family members and familiar locations (B. Bellone, personal communication, September 10, 2021). This practice has a basis in efficiency and ease of use, as AAC applications have become more accessible and the ability to quickly create large vocabularies from pre-populated icon libraries allows speech-language pathologists to serve the needs of more learners with complex communication needs (Elsahar et al., 2019). Given the wide spectrum of individual skills and deficits that may be displayed by learners with complex communication needs (Andzik et al., 2017), this homogenous approach may not meet the needs of every learner. Rather, the best practices in stimulus selection may be a heterogeneous collection of research-informed strategies for helping learners whose communication needs are not met by default teaching practices. One such strategy for facilitating more efficient acquisition and generalization for learners who have not yet maximized their communicative potential could be increasing the visual similarity between AAC icons and their depicted referents.There is evidence that the use of stimuli with greater visual similarity to the target object may lead to more efficient mastery of object-to-picture relations for some individuals with autism (Burren & Dickson, 2019; Goossens', 1983; Hartley & Allen, 2015; Hurlbut et al., 1982). Burren and Dickson (2019) compared acquisition across three different types of pictures: full color photos, black and white line drawings, and black and white stimuli that bore no physical resemblance to their referents (hereafter, arbitrary symbols). For one participant, object-to-picture relations were mastered more efficiently with color photos than with line drawings in eight of 11 comparisons with equivalent mastery rates in the remaining three comparisons, whereas object-to-arbitrary symbol relations were not acquired. These results suggest that the use of stimuli that more closely approximate a photographic depiction of the real-world referent may assist some individuals with complex communication needs in interacting with their verbal communities. An important function of any communication system is to allow the learner to effectively respond to novel stimuli in their environments (Hartley & Allen, 2015), and it is not clear how these results might extend to objects that differ from the target stimulus along one or more dimensions. For example, a full color photo of a red apple may be used to indicate and request a red apple, but will the learner appropriately select the photo of the red apple when given a green apple? A review of the role of picture stimuli in language acquisition for persons with severe cognitive disabilities by Sevcik et al. (1991) put forth a theoretical argument that generalization to novel stimuli may be improved with the use of arbitrary symbols in the graphic request lexicon as compared to stimuli that are more visually similar to their referents. This viewpoint echoed early support for the use of ideographic writing systems; Vanderheiden et al. (1975) predicted that Blissymbols (see Bliss, 1965) would facilitate generalization to novel stimuli because "conceptualization of a thought or idea is not tied to a specific picture." While arguing that greater generalization may be expected with the use of arbitrary symbols, Sevcik et al. acknowledged the experimental evidence that stimuli with more overlapping stimulus features were associated with more efficient acquisition (Goossens', 1983; Hurlbut et al., 1982), corresponding with the results of Burren and Dickson (2019). Such evidence would suggest that the stimuli that yield more rapid acquisition may impede the generalization that would be needed for more functional use of picture-based AAC in real-world settings, where natural stimuli often differ in unpredictable ways. (Abstract shortened by ProQuest). [The dissertation citations contained here are published with the permission of ProQuest LLC. Further reproduction is prohibited without permission. Copies of dissertations may be obtained by Telephone (800) 1-800-521-0600. Web page: http://www.proquest.com.bibliotheek.ehb.be/en-US/products/dissertations/individuals.shtml.]