AI vs academia: Experimental study on AI text detectors' accuracy in behavioral health academic writing.

Artificial Intelligence (AI) language models continue to expand in both access and capability. As these models have evolved, the number of academic journals in medicine and healthcare which have explored policies regarding AI-generated text has increased. The implementation of such policies requires accurate AI detection tools. Inaccurate detectors risk unnecessary penalties for human authors and/or may compromise the effective enforcement of guidelines against AI-generated content. Yet, the accuracy of AI text detection tools in identifying human-written versus AI-generated content has been found to vary across published studies. This experimental study used a sample of behavioral health publications and found problematic false positive and false negative rates from both free and paid AI detection tools. The study assessed 100 research articles from 2016-2018 in behavioral health and psychiatry journals and 200 texts produced by AI chatbots (100 by "ChatGPT" and 100 by "Claude"). The free AI detector showed a median of 27.2% for the proportion of academic text identified as AI-generated, while commercial software Originality.AI demonstrated better performance but still had limitations, especially in detecting texts generated by Claude. These error rates raise doubts about relying on AI detectors to enforce strict policies around AI text generation in behavioral health publications.

Conversational Speech Behaviors Are Context Dependent.

PURPOSE: According to the interpersonal synergy model of spoken dialogue, interlocutors modify their communicative behaviors to meet the contextual demands of a given conversation. Although a growing body of research supports this postulation for linguistic behaviors (e.g., semantics, syntax), little is understood about how this model applies to speech behaviors (e.g., speech rate, pitch). The purpose of this study is to test the hypothesis that interlocutors adjust their speech behaviors across different conversational tasks with different conversational goals. METHOD: In this study, 28 participants each engaged in two different types of conversations (i.e., relational and informational) with two partners (i.e., Partner 1 and Partner 2), yielding a total of 112 conversations. We compared six acoustic measures of participant speech behavior across conversational task and partner. RESULTS: Linear mixed-effects models demonstrated significant differences between speech feature measures in informational and relational conversations. Furthermore, these findings were generally robust across conversations with different partners. CONCLUSIONS: Results suggest that contextual demands influence speech behaviors. These findings provide empirical support for the interpersonal synergy model and highlight important considerations for assessing speech behaviors in individuals with communication disorders.

An exploratory analysis of moderating effects in a non-randomized, quasi-experimental evaluation of a college-based mental health literacy curriculum.

Objective: To assess possible moderating variables that influence the effectiveness of a college-based mental health literacy curriculum, Mental Health Awareness and Advocacy (MHAA). Participants: Students at a Western University in the United States (MHAA treatment group: N = 474; Comparison group: N = 575). Methods: Using a non-randomized, quasi-experimental design over six semesters (2017-2019), students in the treatment and control groups reported on key demographic factors, experiences with mental health, and their knowledge, self-efficacy, and behaviors related to mental health literacy in a pretest/post-test format. Linear mixed effects models were used to assess moderation. Results: MHAA curriculum's efficacy varied based on participants' current depression and anxiety levels, whether they had prior personal experience with a mental health concern, whether they had been diagnosed with a mental health concern, and whether they had experienced suicidal thoughts. Conclusions: Reducing anxiety and depression among students may increase the efficacy of the MHAA program.

Lexical Alignment is Pervasive Across Contexts in Non-WEIRD Adult-Child Interactions.

Lexical alignment, a communication phenomenon where conversational partners adapt their word choices to become more similar, plays an important role in the development of language and social communication skills. While this has been studied extensively in the conversations of preschool-aged children and their parents in Western, Educated, Industrialized, Rich, and Democratic (WEIRD) communities, research in other pediatric populations is sparse. This study makes significant expansions on the existing literature by focusing on alignment in naturalistic conversations of school-aged children from a non-WEIRD population across multiple conversational tasks and with different types of adult partners. Typically developing children aged 5 to 8 years (n = 45) engaged in four semi-structured conversations that differed by task (problem-solving vs. play-based) and by partner (parent vs. university student), resulting in a corpus of 180 conversations. Lexical alignment scores were calculated and compared to sham conversations, representing alignment occurring at the level of chance. Both children and adults coordinated their conversational utterances by re-using or aligning each other's word choices. This alignment behavior persisted across conversational tasks and partners, although the degree of alignment was moderated by the conversational context. These findings suggest that lexical alignment is a robust phenomenon in conversations between school-age children and adults. Furthermore, this study extends lexical alignment findings to a non-WEIRD culture, suggesting that alignment may be a coordination strategy employed by adults and children across diverse linguistic and cultural backgrounds.

An ecological examination of early adolescent e-cigarette use: A machine learning approach to understanding a health epidemic.

E-cigarette use among adolescents is a national health epidemic spreading faster than researchers can amass evidence for risk and protective factors and long-term consequences associated with use. New technologies, such as machine learning, may assist prevention programs in identifying at risk youth and potential targets for intervention before adolescents enter developmental periods where e-cigarette use escalates. The present study utilized machine learning algorithms to explore a wide array of individual and socioecological variables in relation to patterns of lifetime e-cigarette use during early adolescence (i.e., exclusive, or with tobacco cigarettes). Extant data was used from 14,346 middle school students (Mage = 12.5, SD = 1.1; 6th and 8th grades) who participated in the Utah Prevention Needs Assessment. Students self-reported their substance use behaviors and related risk and protective factors. Machine learning algorithms examined 112 individual and socioecological factors as potential classifiers of lifetime e-cigarette use outcomes. The elastic net algorithm achieved outstanding classification for lifetime exclusive (AUC = .926) and dual use (AUC = .944) on a validation test set. Six high value classifiers were identified that varied in importance by outcome: Lifetime alcohol or marijuana use, perception of e-cigarette availability and risk, school suspension(s), and perceived risk of smoking marijuana regularly. Specific classifiers were important for lifetime exclusive (parent's attitudes regarding student vaping, best friend[s] tried alcohol or marijuana) and dual use (best friend[s] smoked cigarettes, lifetime inhalant use). Our findings provide specific targets for the adaptation of existing substance use prevention programs to address early adolescent e-cigarette use.

Dimensions of caregiver strain may partially mediate the relationship between youth symptomology and counseling utilization among Latinxs.

Whether Latinx families use youth mental health services (MHS) depends on complex influences of barriers and facilitators within and outside of the home. This research sought to shed light on caregiver strain as part of the equation focused on parental identification and responses to youth mental health needs. We examined multiple dimensions of caregiver strain as potential mediators between youth mental health symptom severity and psychological counseling utilization. The present sample consisted of 598 Latinx caregivers to youths ages 6-18 who provided information on youth internalizing and externalizing problems, caregiver strain, and youth psychological counseling service utilization within the last year. Our findings suggest that youth symptom severity (internalizing and externalizing problems) was generally positively associated with dimensions of caregiver strain. Youth symptom severity through objective and subjective internalized strain pathways were associated with greater odds of youth MHS utilization. In contrast, youth symptom severity through subjective externalized strain reduced the odds that Latinx caregivers would report utilizing youth MHS. These models only partially mediated the relationship between youth problems and service use. Findings suggest that Latinx caregivers may navigate conflicting sources of strain related to their child's mental health problem severity in ways that may differentially impact the odds that they access youth MHS. Along with addressing structural and systemic barriers to care, utilization of psychological counseling services may also be improved through interventions that help Latinx caregivers view youth services as avenues for addressing caregiver strain and providing psychoeducation that frames externalized strain within a mental health lens.