Exploratory Results From a Chatbot-Based Screening Tool for Computer-Mediated Communication Impairment

Purpose: Computer-mediated communication (CMC) is a routine language modality, yet its disorder profile is poorly understood. Guided by exploratory frameworks, this study designed SpeechBot, a brief chatbot-based screener. It was piloted in an acute-care hospital to evaluate the tool's clinical feasibility and to probe for preliminary patterns of CMC impairment. Method: Thirty-three adults who reported regularly texting and could complete three inclusion items were screened during routine bedside speech and language evaluations. Speech-language pathologists summarized function across traditional language domains on 5-point Likert scales, while a CMC impairment score (CMCIS; 0–12) captured CMC function. Consistent with exploratory aims, no confirmatory hypotheses were tested. Results: Impairment scores clustered into three strata: no–minimal (CMCIS 0–3; 79%), moderate (4–6; 15%), and severe (≥ 7; 6%). Seven participants with moderate-to-severe CMCIS scores demonstrated no observed deficits in traditional speech-language domains during bedside evaluations. Higher CMCIS was associated with slower response times (ρ = .56, p = .001), but not with age or clinician-rated language scores. Conclusion: These exploratory findings support the feasibility of chatbot-based CMC screening and generate several lines of future inquiry, including refining CMC impairment metrics; conducting rigorous experimental research to contextualize CMC impairment against established cognitive–linguistic deficits; and validating CMC screening in larger, demographically diverse samples.