Open research questions in Mental Health Research Topics

What the literature leaves open

• ABSTRACT Short‐form video addiction (SFVA) is increasingly prevalent among adolescents and closely related to time perspective, yet their associations at the symptom and dimensional levels, particularly gender differences, remain underexplored.

  Gender‐Specific Networks in the Relationship Between Time Perspective and Short‐Form Video Addiction Among Chinese Adolescents · 2026 · DOI

• Our findings suggest that self-related ability beliefs in a performance context can be shaped by two distinct, yet concurrently occurring biases.

  Causal attributions shape the formation of novel ability self-beliefs · 2026 · DOI

• While the present findings provide initial insights into the role of causal attributions in the process of forming new ability self-beliefs, several con- siderations are important for their interpretation. First, the sample size may limit statistical power and the robustness of the observed effects. A sensi- tivity analysis indicated that the present design was primarily powered to detect medium-sized associations, suggesting that smaller effects may not have been reliably detected. Accordingly, the observed associations between self-esteem and attributional biases should be interpreted with some caution and warrant further investigation in larger samples to establish their stability and generalizability. Second, given that the study was conducted in a student sample, depressive symptom levels were likely to fall predominantly below clinical thresholds, and participants’ diagnostic status was not assessed. As a result, effects involving depressive symptoms may be underestimated due to limited between-subject variability in PHQ-9 scores. Stronger or qualita- tively different effects might emerge in samples that include individuals with a clinical diagnosis of depression. Third, given that depressive symptoms and self-esteem were not experimentally manipulated, the directionality and causality of their associations with attributional patterns remain unclear. Theories assume recursive dynamics between aspects of the self-concept and attributional patterns26,27,52–54. While our current analysis demonstrates self-serving attributional patterns and shows how these may influence belief updating over time, it does not allow conclusions about the causal interplay between the resulting beliefs and self-esteem or depressive symptoms, nor about whether such beliefs subsequently shape future attributions. Again, the relatively small sample size may additionally constrain the ability to test more complex reciprocal or directional effects. Despite these issues, the present findings contribute preliminary evidence that is consistent with contemporary psychological theories26,27 and previous findings in related fields28,29,31, and may help to set a direction for future studies examining attributional patterns and effects on self-belief formation.

  Causal attributions shape the formation of novel ability self-beliefs · 2026 · DOI

• Several limitations should be acknowledged. First, the sample was drawn from the general population, and network structures may differ in clinical samples (Robinaugh et al., 2020). Second, NIRA ranks components based on their overall impact on the total network state and does not distinguish between within- and cross-cluster propa- gation. Because the anxiety cluster contains more nodes and stronger internal con- nectivity than the life-domain cluster, symptom perturbations are structurally more likely to produce larger system-level effects, particularly in aggravating simulations; therefore, lower rankings of life-domain components should not be interpreted as evidence of weak influence. Third, as simulations were based on cross-sectional data, future studies using intensive longitudinal designs are needed to validate these findings and better capture dynamic psychological processes (Haslbeck et al., 2022; Henry et al., 2022). Fourth, it should also be noted that cross-sectional network mod- els do not permit causal inference. Edges represent conditional associations rather than directional effects, and centrality indices should not be interpreted as definitive targets for intervention (Neal et al., 2022). Moreover, network parameters may be sensitive to estimation choices and sample characteristics. Although bootstrap pro- cedures were used to assess stability, replication across independent samples is nec- essary to confirm the robustness of these findings. Fifth, because the present study conceptualizes life domains as components within a structured vulnerability con- figuration, future research should also examine whether similar structural patterns emerge under different cultural and educational systems. The present findings are based on a large sample of Chinese youth, and cultural context may shape both levels of DLS and their associations with anxiety symptoms. Cultural norms emphasiz- ing academic achievement, social evaluation, and interdependence may amplify the 1 3Age Differences in the Interconnections between Domain-Specific Life… salience of self- and school-related domains in the Chinese context. Consequently, the relative importance and network structure of life domains may differ across soci- eties with distinct cultural values or educational systems. Future research using cross- cultural samples is needed to examine the generalizability of these network patterns and to identify culturally invariant versus culture-specific mechanisms linking life satisfaction and anxiety (Schimmack et al., 2005). Sixth, the present study focused on associations between domain-specific life satisfaction and anxiety symptoms. Other internalizing problems, such as depressive symptoms or psychological distress, may exhibit different network configurations. Future research should examine whether similar domain-specific vulnerability patterns emerge for other forms of internal- izing psychopathology. Seventh, although GAD-7 showed measurement invariance across age groups, domain-specific life satisfaction was assessed using single-item measures. Future research may further examine potential developmental differences in measurement functioning using multi-item assessments.

  Age Differences in the Interconnections between Domain-Specific Life Satisfaction and Anxiety Symptoms Across Adolescence and Emerging Adulthood: Network-Based Analyses · 2026 · DOI

• There are many strengths of the study and EMA approaches in general. EMA offers the opportunity to measure daily cognitive functioning in naturalistic environments, which https://mhealth.jmir.org/2026/1/e64397 can enhance our understanding of contemporaneous, WP relationships between cognitive and affective processes in clinical and community samples. In clinical assessments of cognition, there may be nuances that emerge outside the JMIR Mhealth Uhealth 2026 | vol. 14 | e64397 | p.

  Daily Negative Affect and Reaction Time Inconsistency in Emerging Adults: Ecological Momentary Assessment Study. · 2026 · DOI

• of the GMV platform, together with the randomized design of two-stage informed study, questionnaire combinations. the development of informed the the The email invitation contained two response links (“Yes”/“No”), each directing recipients to a corresponding questionnaire. Those selecting “Yes” completed a brief sociodemographic questionnaire capturing demographic characteristics and indicators of study interest; those selecting “No” completed a brief questionnaire capturing reasons for non-participation and incentive preferences (see Supplementary File S1). to created instances were To ensure responses were correctly attributed the participant’s assigned incentive condition and user category, separate questionnaire each incentive condition (financial, social, nocombination of incentive), response option (Yes/No), and user category (residents, visitors, commercial occupiers), yielding 18 instances in total. Each instance was linked via a condition-specific URL and labeled with a structured identifier to support tracking and analysis (e.g., RFNY=Resident–Financial Incentive–Yes; VNICN=Visitor–No Incentive–No). for 2.11. Data management Several procedures were implemented to uphold data quality and ensure data security. The GMV survey design required responses for four core demographic items (gender, ethnic group, age range, and relationship to King’s Cross), while follow-up preference items were optional. Additionally, the platform’s device recognition feature guaranteed that once a participant submitted a response, they were unable to access the survey again, the uniqueness of each submission. As data were collected directly in a pre-coded digital format on the GMV platform, manual data entry was unnecessary, reducing the risk of transcription errors and enhancing the overall accuracy of the dataset. thereby preserving The primary outcome—the participant’s binary decision to participate (“Yes” or “No”)—was derived directly from the raw data exported from the GMV platform. Each survey response was coded post-collection: responses from participants who completed the “Yes” survey were assigned a value of 1 and those who completed the “No” survey were assigned a value of 0. This coding process ensured consistent and accurate representation of the primary outcome for analysis. All data storage, handling, and sharing procedures will comply with institutional data security and ethical guidelines. Study data will be cleaned and anonymised prior to any sharing. Access to identifiable and cleaned study data will be restricted to authorised members of the research team and approved project partners in accordance with prior datasharing agreements. 2.12. Analysis plan for reporting The trial will be reported in accordance with the CONSORT trials. The detailed randomized guidelines statistical analysis plan has been published elsewhere (34). All analyses will follow the intention-to-treat principle, including all randomized invitees, to ensure a comprehensive assessment of the intervention’s effectiveness (35). The primary analysis will compare the effectiveness of the three intervention arms. Secondary analyses will explore subgroup differences in response to the incentives as well as examine barriers to participation and alternative preferences.

  Examining the effectiveness of financial vs. social incentives to participate in a smartphone ecological momentary assessment well-being study: protocol for a randomized controlled trial · 2026 · DOI

• rather than causal First, the cross-sectional design precludes causal inference; central or bridge nodes should be interpreted as correlational signals leverage points. These findings are hypothesis-generating, and claims about mechanisms or intervention relevance require longitudinal or experimental evidence. Second, the reliance on self-reported measures may introduce biases. The IPAQ-SF may not be appropriate for the youngest participants, and measurement invariance of the GAD-7 across school stages was not formally tested. Sleep and dietary behaviors were assessed using brief self-developed items. Supplementary item-level analyses suggested that the central role of MPA was preserved, whereas the constituent sleep- and dietrelated items showed heterogeneous contributions, particularly for sleep. Future research should employ validated instruments such as the Pittsburgh Sleep Quality Index.

  Network analysis of lifestyle behaviors and anxiety in children and adolescents: gender and school-stage heterogeneity · 2026 · DOI

• Our analyses highlight critical limitations in the use of SCF summary statistics (e.g., means, variance, and standard deviations) within the context of EMA studies where the assumption of equal numbers of observations per subject rarely holds, missing data are common, and ordinal scales with fewer response options are frequently used. The SCF method, while traditionally employed, falls short in several key areas when faced with these inherent challenges of EMA data. Specifically, it inaccurately assumes equal precision across subjects with varying data sizes, does not consider correlated nature of within-subject data, treats summary statistics as fixed rather than estimated quantities with error, and operates under the assumption that missing responses occur completely at random (MCAR). When such summary statistics are used as predictors in stage-2 regression models, this introduces bias in the estimation of regression coefficients, as evidenced by our simulation studies, but also skewed standard errors for the stage 2 regressors, thereby inflating the risk of Type-I errors. In contrast, the MELS model offers a robust alternative that helps to address many of these issues. This approach can adjust for differences in the number of observations per subject, recognizing unique random scale across subjects and the correlation within subjects' data, thereby providing more accurate estimates. Moreover, the incorporation of plausible value replications acknowledges and accounts for the inherent uncertainty in estimation, a critical consideration overlooked by the SCF method. Most importantly, by employing full maximum likelihood estimation, the MELS model operates within the MAR assumption for handling missing data (Walters, 2015). This assumption is much more realistic for EMA data than the MCAR assumption, offering a significant improvement over traditional SCF approach. Overall, current evidence supports the MELS model as a superior analytical strategy for Ecological Momentary Assessment (EMA) data, particularly in studies with unbalanced designs and missing data. The MELS approach yields more reliable and valid inferences by enhancing the precision of estimates and accommodating the complexities inherent in real-world EMA data collection, such as ordinal scale responses and varying mean-to-variance ratios. This represents a significant methodological advancement in behavioral and biostatistics research. The accessibility of the MELS model has also increased substantially, with implementation now available in major statistical packages including SAS (PROC NLMIXED) and R (fastregls) (Gill & Hedeker, 2023; Ma & Hedeker, 2020). Additionally, the MixWILD software provides a free, open-source, and userfriendly desktop GUI specifically designed for the analysis of MELS models in intensive longitudinal data (Dzubur et al., 2020). As the adoption of the MELS approach becomes increasingly feasible, it is crucial to apply this methodology to more nuanced data settings, thereby advancing knowledge in behavioral science. Journal of Behavioral Medicine1 3 Appendix Simulated data—SAS code (missing completely at random (MCAR)) The provided SAS code is designed to simulate a multilevel dataset with a Missing Completely at Random (MCAR) mechanism, which is useful for evaluating statistical methods under controlled missing data conditions. Several key elements in the code deserve attention: ● The simulation creates 500 datasets (ndats = 500), each containing 300 subjects (ns = 300). This ensures a sufficiently large sample size to study the statistical properties of the MCAR mechanism and its impact on analysis. ● Each subject has 30 repeated observations (nis = 30).

  Modeling intraindividual means and variances from ecological momentary assessment data: comparing standard computational formulas to mixed-effects location-scale model estimates · 2026 · DOI

• The identified modular structure (sleep disorder-dryness-fatigue clustering in postmenopausal stage) suggests co-occurring symptoms may share mechanisms, but no longitudinal follow-up data were collected to test whether bundled nonpharmacological interventions targeting symptom clusters are more effective than isolated symptom treatment in climacteric syndrome management.

  Network analysis of patient-reported complaints in climacteric syndrome: a single-center study · 2026 · DOI

• The study relied on unstructured chief complaint data from electronic medical records rather than validated instruments (MENQOL, MRS, PHQ), which revealed pain as a central symptom unlike previous network analyses. Direct comparison studies combining both unstructured clinical complaint networks and structured questionnaire-based symptom networks for climacteric syndrome are needed to determine whether symptom centrality findings are measurement-dependent.

  Network analysis of patient-reported complaints in climacteric syndrome: a single-center study · 2026 · DOI

• Potential omissions and categorization errors may have occurred during symptom extraction and rephrasing from unstructured medical records in the climacteric syndrome analysis. Future network studies should employ structured symptom extraction protocols with inter-rater reliability validation and natural language processing algorithms to minimize data entry bias and ensure symptom network structure reproducibility.

  Network analysis of patient-reported complaints in climacteric syndrome: a single-center study · 2026 · DOI

• The study identified distinct symptom clusters in premenopausal and postmenopausal networks but did not empirically test the underlying mechanisms linking these clusters. Future research should investigate the shared physiological pathways and hormonal mechanisms connecting sleep disorder-dryness-fatigue in postmenopausal networks and headache-dizziness-palpitation in premenopausal networks using biomarker analysis alongside network analysis.

  Network analysis of patient-reported complaints in climacteric syndrome: a single-center study · 2026 · DOI

• The symptom network structure was derived exclusively from a single Korean Medicine hospital, limiting generalizability across diverse clinical settings. Multicenter network analysis studies of climacteric syndrome symptoms across general hospitals, primary care clinics, gynecology clinics, and community-based settings in different countries are needed to validate whether the identified central symptoms (pain, hot flush) and modular patterns remain consistent across cultural and ethnic populations.

  Network analysis of patient-reported complaints in climacteric syndrome: a single-center study · 2026 · DOI

• The study was unable to analyze correlations among age, duration since menopause, and hormone therapy use due to small sample size and limited data. Future network analysis studies on climacteric syndrome should incorporate larger datasets with detailed demographic and therapeutic variables to examine how these factors influence symptom network structure and centrality patterns.

  Network analysis of patient-reported complaints in climacteric syndrome: a single-center study · 2026 · DOI

• Reliance on self-reported GDS-15 data may be problematic in cognitively impaired individuals or those with subclinical delirium despite exclusion of overt delirium.

  Depressive symptoms in hospitalized geriatric patients with and without cognitive impairment: a cross-sectional network analysis approach · 2026 · DOI

• Exclusion of individuals with a MMSE score of 16 points or less may have introduced selection bias and restricts generalizability to those with more severe cognitive deficits.

  Depressive symptoms in hospitalized geriatric patients with and without cognitive impairment: a cross-sectional network analysis approach · 2026 · DOI

• Furthermore, measurement invariance between different treatment groups could not be established, strongly indicating that not all respondents use the PHQ-9 in the same manner as the only differentiating factor between groups was their randomized treatment group allocation.

  Why are we still using the PHQ-9? A Historical Review and Psychometric Evaluation of Measurement Invariance · 2025 · DOI

• The Affective Reactivity Index (ARI) is one of the most studied scales for assessing youth irritability, but little is known about its measurement performance in community populations.

  Psychometric Evaluation of the Affective Reactivity Index Among Children and Adolescents in China: A Multi-Method Assessment Approach · 2023 · DOI

• Integration of advances in psychiatric genetics, neurobiological computation, and data science into a unified etiological and pathophysiological framework requires systematic identification of the missing scientific pieces concerning genetic, developmental, and environmental contributors to psychiatric disorder mechanisms.

  Psychiatric Disorders: Grounded in Human Biology but Not Natural Kinds · 2021 · DOI

• The stability and instability of individual symptoms and diagnosable psychiatric disorders across the developmental period between childhood and young adulthood requires longitudinal investigation to establish when and how diagnostic thresholds should be applied clinically.

  Psychiatric Disorders: Grounded in Human Biology but Not Natural Kinds · 2021 · DOI

• Case-control research designs that exclude or ignore comorbidity must be systematically replaced with cohort selection methods that capture the continuous distribution of etiological factors and symptoms in psychiatric populations to align research practices with dimensional nosology.

  Psychiatric Disorders: Grounded in Human Biology but Not Natural Kinds · 2021 · DOI

• The genetic heterogeneity of psychiatric disorders necessitates systematic studies of how reassortment of risk alleles during meiosis generates new combinations and degrees of genetic loading across generations, particularly to explain the emergence of spectrum phenotypes and intermediate presentations in families with schizophrenia loading.

  Psychiatric Disorders: Grounded in Human Biology but Not Natural Kinds · 2021 · DOI

• Research Domain Criteria (RDoC) requires expanded investigation of the relationship between specific brain circuits and individual symptoms across genetics, normal and pathologic brain biology, and computational models of cognition and behavior to validate its dimensional approach to psychiatric nosology.

  Psychiatric Disorders: Grounded in Human Biology but Not Natural Kinds · 2021 · DOI

• HiTOP would benefit from systematic collection of symptom-level reporting across genetic, clinical, and epidemiological studies to facilitate deconstruction of DSM disorders and increase available information for symptom covariation analyses within its hierarchical framework.

  Psychiatric Disorders: Grounded in Human Biology but Not Natural Kinds · 2021 · DOI

• Emerging evidence suggests impulsive states may be reliably measured in the moment using ecological momentary assessment (EMA); however, research has not investigated whether the multi-factor structure of impulsive traits also characterizes impulsive states.

  Impulsive States and Impulsive Traits: A Study of the Multilevel Structure and Validity of a Multifaceted Measure of Impulsive States · 2020 · DOI

Most-cited papers in Mental Health Research Topics

• A Self-Rating Depression Scale · Archives of General Psychiatry · 1965 · 7,759 citations
• The relation between short-term emotion dynamics and psychological well-being: A meta-analysis. · Psychological Bulletin · 2015 · 815 citations
• Supervised Machine Learning: A Brief Primer · Behavior Therapy · 2020 · 749 citations
• Reflecting on rumination: Consequences, causes, mechanisms and treatment of rumination · Behaviour Research and Therapy · 2020 · 720 citations
• Testing prospective effects in longitudinal research: Comparing seven competing cross-lagged models. · Journal of Personality and Social Psychology · 2020 · 639 citations
• Best practice guidance for linear mixed-effects models in psychological science · Journal of Memory and Language · 2020 · 604 citations
• The Interrelationship of Neuroticism, Sex, and Stressful Life Events in the Prediction of Episodes of Major Depression · American Journal of Psychiatry · 2004 · 599 citations
• CINeMA: Software for semiautomated assessment of the confidence in the results of network meta‐analysis · Campbell Systematic Reviews · 2020 · 457 citations
• Predicting depressive symptoms with a new measure of shame: The Experience of Shame Scale · British Journal of Clinical Psychology · 2002 · 445 citations
• Clinical implications of "subthreshold" depressive symptoms. · Journal of Abnormal Psychology · 2000 · 395 citations

Most recent work

• The Human Experience Framework: A Neuroscience-Grounded Clinical Manual for Body-Based Therapeutic Assessment and Intervention · Open MIND · 2026
• The Association Between Childhood Adversity and Cognitive Aging Trajectory in Later Life: Evidence from the China Health and Retirement Longitudinal Study (CHARLS) · Clinical Gerontologist · 2026
• Frustration of basic psychological needs as a potential transdiagnostic pathway between anger suppression, self-blame, and depressive symptoms · Journal of Social and Clinical Psychology · 2026
• Identifying symptom communities and core symptoms in the anxiety-depression network among computer science students · Scientific Reports · 2026
• Unraveling links between lifestyle behaviors and depressive symptoms: a network analysis using NHANES 2007–2018 data · Annals of General Psychiatry · 2026
• Generative Artificial Intelligence Models for Developing Neuroimaging Markers of Psychiatric Disorders · Biological Psychiatry · 2026
• Developing Clinically Useful Applications of Network Analysis in Mood Disorders Psychiatry · Psychiatric Annals · 2026
• Network structure, invariance, and symptom communities of anxiety-depression network among college student-athletes: a network analysis · International Journal of Sport and Exercise Psychology · 2026
• Unveiling the complex interactions of mindset, emotions, and self-regulated learning in EFL writing: A latent profile and network analysis · Assessing Writing · 2026
• Mapping the structure of student burnout in online learning: An integrated Gaussian model and directed acyclic graph approach · The Internet and Higher Education · 2026

Related topics in Psychology