baseline systems on the three proposed tasks: state-of-mind recognition, depression assessment with AI, and cross-cultural affect sensing, respectively.
The INTERSPEECH 2020 Computational Paralinguistics Challenge addresses three different problems for the first time in a research competition under well-defined conditions: In the Elderly Emotion Sub-Challenge, arousal and valence in the speech of elderly individuals have to be modelled as a 3-class problem; in the Breathing Sub-Challenge, breathing has to be assessed as a regression problem; and in the Mask Sub-Challenge, speech without and with a surgical mask has to be told apart. We describe the Sub-Challenges, baseline feature extraction, and classifiers based on the 'usual' COMPARE and BoAW features as well as deep unsupervised representation learning using the AUDEEP toolkit, and deep feature extraction from pre-trained CNNs using the DEEP SPECTRUM toolkit; in addition, we partially add deep end-to-end sequential modelling, and, for the first time in the challenge, linguistic analysis.
Background Mobile health apps (MHA) have the potential to improve health care. The commercial MHA market is rapidly growing, but the content and quality of available MHA are unknown. Instruments for the assessment of the quality and content of MHA are highly needed. The Mobile Application Rating Scale (MARS) is one of the most widely used tools to evaluate the quality of MHA. Only few validation studies investigated its metric quality. No study has evaluated the construct validity and concurrent validity. Objective This study evaluates the construct validity, concurrent validity, reliability, and objectivity, of the MARS. Methods Data was pooled from 15 international app quality reviews to evaluate the metric properties of the MARS. The MARS measures app quality across four dimensions: engagement, functionality, aesthetics and information quality. Construct validity was evaluated by assessing related competing confirmatory models by confirmatory factor analysis (CFA). Non-centrality (RMSEA), incremental (CFI, TLI) and residual (SRMR) fit indices were used to evaluate the goodness of fit. As a measure of concurrent validity, the correlations to another quality assessment tool (ENLIGHT) were investigated. Reliability was determined using Omega. Objectivity was assessed by intra-class correlation. Results In total, MARS ratings from 1,299 MHA covering 15 different health domains were included. Confirmatory factor analysis confirmed a bifactor model with a general factor and a factor for each dimension (RMSEA = 0.074, TLI = 0.922, CFI = 0.940, SRMR = 0.059). Reliability was good to excellent (Omega 0.79 to 0.93). Objectivity was high (ICC = 0.82). MARS correlated with ENLIGHT (ps<.05). Conclusion The metric evaluation of the MARS demonstrated its suitability for the quality assessment. As such, the MARS could be used to make the quality of MHA transparent to health care stakeholders and patients. Future studies could extend the present findings by investigating the re-test reliability and predictive validity of the MARS.
Background: Eating problems are highly prevalent among young adults. Universities could be an optimal setting to prevent the onset of eating disorders through psychological intervention. As part of the World Mental Health-International College Student initiative, this systematic review and meta-analysis synthesizes data on the efficacy of eating disorder prevention programs targeting university students.Method: A systematic literature search of bibliographical databases (CENTRAL, MEDLINE, PsycINFO) for randomized trials comparing psychological preventive interventions for eating disorders targeting university students with psychoeducation or inactive controls was performed on October 22, 2019.Results: Twenty-seven studies were included. Thirteen (48.1%) were rated to have a low risk of bias. The relative risk of developing a subthreshold or full-blown eating disorder was incidence rate ratio = 0.62 (95% CI [0.44, 0.87], n c = 8, numbersneeded-to-treat [NNT] = 26.08; standardized clinical interviews only), indicating a 38% decrease in incidence in the intervention groups compared to controls. Small to moderate between-group effects at posttest were found on eating disorder symptoms (g = 0.35, 95% CI [0.24, 0.46], NNT = 5.10, n c = 26), dieting (g = 0.43, 95% CI [0.29, 0.57], NNT = 4.17, n c = 21), body dissatisfaction (g = 0.40, 95% CI [0.27, 0.53], NNT = 4.48, n c = 25), drive for thinness (g = 0.43, 95% CI [0.27, 0.59], NNT = 4.23, Int J Eat Disord. 2020;53:813-833.wileyonlinelibrary.com/journal/eat 813 n c = 12), weight concerns (g = 0.33, 95% CI [0.10, 0.57], NNT = 5.35, n c = 13), and affective symptoms (g = 0.27, 95% CI [0.15, 0.38], NNT = 6.70, n c = 18). The effects on bulimia nervosa symptoms were not significant. Heterogeneity was moderate across comparisons.Discussion: Eating disorder prevention on campus can have significant, small-tomoderate effects on eating disorder symptoms and risk factors. Results also suggest that the prevention of subthreshold and full-syndrome eating disorders is feasible using such interventions. More research is needed to identify ways to motivate students to use preventive eating disorder interventions. AbstractAntecedentes: Los trastornos de la conducta alimentaria son altamente prevalentes entre los adultos jóvenes. Las universidades podrían ser un entorno óptimo para prevenir la aparición de trastornos alimentarios a través de la intervención psicológica. Como parte de la iniciativa World Mental Health-International College Student, esta revisión sistemática y meta-análisis sintetiza datos sobre la eficacia de los programas de prevención de trastornos alimentarios dirigidos a estudiantes universitarios. Método: Una búsqueda bibliográfica sistemática de datos bibliográficas (CENTRAL, MEDLINE, PsycINFO) para ensayos aleatorios que comparaban intervenciones preventivas psicológicas para trastornos alimentarios dirigidos a estudiantes universitarios con psicoeducación o controles inactivos fue realizada hasta el 22 de octubre de 2019. Resultados: Se incluyeron 27 estudios. T...
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.