arxiv: 2601.07961 · v2 · submitted 2026-01-12 · 💻 cs.CL · stat.AP

Recognition: 1 theorem link

· Lean Theorem

Language Markers of Emotion Flexibility Predict Depression and Anxiety Treatment Outcomes

Benjamin Brindle , George A. Bonanno , Thomas Derrick Hull , Nicolas Charon , Matteo Malgaroli

Authors on Pith no claims yet

Pith reviewed 2026-05-16 14:36 UTC · model grok-4.3

classification 💻 cs.CL stat.AP

keywords emotional flexibilitytreatment outcomesanxietydepressionlinguistic markersteletherapystate-space modelemotion dynamics

0 comments

The pith

Emotion dynamics from therapy transcripts predict which anxiety and depression patients will fail to improve.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The study processes 12 weeks of de-identified teletherapy transcripts from 12,043 U.S. patients with moderate-to-severe anxiety and depression. A transformer model extracts emotions at each talk turn, and a state-space model clusters the resulting sequences into an improving group of 8,230 patients and a non-response group of 3,813 patients. The non-response group shows higher odds of symptom worsening and lower rates of clinically significant improvement, with sadness and fear dominating their emotion networks. Improving patients instead display more balanced patterns that include joy and neutral states. These linguistic markers of emotional inflexibility therefore supply a passive, scalable signal for identifying patients at elevated risk of poor treatment response.

Core claim

Transformer-based extraction of emotions at the talk-turn level from teletherapy transcripts, followed by state-space clustering of their temporal dynamics, partitions 12,043 patients into an improving subgroup and a non-response subgroup; the latter exhibits elevated odds of deterioration, reduced likelihood of clinically significant change, and emotion networks in which sadness and fear exert the strongest influence, while improving patients maintain balanced joy, sadness, and neutral expressions.

What carries the argument

Transformer-based small language model for turn-level emotion extraction combined with VISTA-SSM state-space clustering that produces temporal emotion networks.

If this is right

Linguistic markers of emotional inflexibility can stratify patients by risk of non-response before full treatment completion.
Non-responders are characterized by sadness and fear as the dominant drivers of emotion dynamics.
Improving patients maintain more balanced emotion networks that include joy and neutral states.
The approach supplies a scalable, passive method for treatment risk stratification using existing session transcripts.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Continuous monitoring of emotion flexibility during sessions could allow therapists to adjust interventions in real time.
The same transcript-based clustering could be tested for predictive value in other emotion-regulation disorders such as PTSD.
If the signal holds, routine outcome tracking could shift from infrequent questionnaires to automated analysis of every session.
Targeted exercises that increase emotional variety might be evaluated as an add-on to standard treatment for patients flagged by the model.

Load-bearing premise

The emotion labels extracted by the transformer model accurately reflect patients' actual emotional states and the resulting clusters capture real predictive dynamics rather than artifacts of the model or data.

What would settle it

An independent validation study that rates a subset of transcripts by blinded clinicians or patients themselves and then tests whether the original group assignments still predict measured clinical improvement at 12 weeks.

read the original abstract

Predicting treatment non-response for anxiety and depression is challenging, in part because of sparse symptom assessments in real-world care. We examined whether passively captured, fine-grained emotions serve as linguistic markers of treatment outcomes by analyzing 12 weeks of de-identified teletherapy transcripts from 12,043 U.S. patients with moderate-to-severe anxiety and depression symptoms. A transformer-based small language model extracted patients' emotions at the talk-turn level; a state-space model (VISTA-SSM) clustered subgroups based on emotion dynamics over time and produced temporal networks. Two groups emerged: an improving group (n=8,230) and a non-response group (n=3,813) showing increased odds of symptom deterioration, and lower likelihood of clinically significant improvement. Temporal networks indicated that sadness and fear exerted most influence on emotion dynamics in non-responders, whereas improving patients showed balanced joy, sadness, and neutral expressions. Findings suggest that linguistic markers of emotional inflexibility can serve as scalable, interpretable, and theoretically grounded indicators for treatment risk stratification.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper uses turn-level emotion extraction and state-space clustering on a large therapy transcript set to flag non-responders, but the lack of reported validation on the emotion labels leaves the main claims under-supported.

read the letter

The core finding is that a transformer pulled emotions from 12,043 teletherapy transcripts, a VISTA-SSM model split patients into an improving group and a non-response group, and the non-responders showed higher odds of worsening symptoms with sadness and fear dominating their temporal networks. The scale is the real strength here: real-world data over 12 weeks, linked to actual clinical outcomes, and a dynamic rather than static view of emotion patterns. That moves past the usual small-sample linguistic marker studies and gives a concrete pipeline for passive monitoring. What works is the attempt to ground the clusters in observable symptom change instead of just internal model fit. The soft spots are straightforward. No accuracy numbers, F1 scores, or even basic checks on how well the small language model labels clinical talk are given, so any noise in sadness or fear detection would flow straight into the clusters and the reported network influences. The abstract also skips baseline comparisons and details on how the unsupervised clustering was tuned or tested for stability, which makes it hard to judge whether the two-group split is robust or just an artifact of the chosen parameters. This is worth a look for anyone building tools for digital mental health or computational approaches to therapy monitoring. Readers who care about scalable prediction in real care settings will find the setup useful even if they have to treat the specific group differences as preliminary. I would send it for peer review because the dataset size and the outcome linkage give it enough substance to justify referee time, though the authors will need to add validation metrics and robustness checks before the results can be taken at face value.

Referee Report

3 major / 2 minor

Summary. The paper analyzes 12,043 de-identified teletherapy transcripts from U.S. patients with moderate-to-severe anxiety and depression. A transformer-based small language model extracts emotions at the talk-turn level; VISTA-SSM clusters patients into an improving group (n=8,230) and a non-response group (n=3,813) based on emotion dynamics, then constructs temporal networks. The non-response group shows higher odds of symptom deterioration and lower likelihood of clinically significant improvement, with sadness and fear exerting greater influence in their networks versus balanced expressions in improvers. The central claim is that linguistic markers of emotional inflexibility can serve as scalable predictors for treatment risk stratification.

Significance. If the core associations hold after validation, the work offers a passive, scalable approach to risk stratification in real-world mental health care using fine-grained linguistic data from routine sessions. The large sample and state-space modeling of dynamics are notable strengths that could support theoretically grounded, interpretable indicators beyond sparse symptom assessments.

major comments (3)

[Abstract/Methods] Abstract and Methods: No accuracy, F1, or inter-annotator metrics are reported for the transformer-based small language model on clinical therapy transcripts. This is load-bearing, as misclassification rates above ~20-30% for key emotions (sadness, fear) would propagate directly into the VISTA-SSM clusters, temporal networks, and the reported outcome associations for the n=3,813 non-response group.
[Results] Results: The manuscript provides no baseline comparisons, error bars, confidence intervals, or details on how symptom outcome measures were statistically linked to the emotion-derived clusters. Without these, the claims of increased deterioration odds and lower improvement likelihood in the non-response group cannot be evaluated for robustness.
[Methods] Methods (VISTA-SSM): Clustering and network construction are performed on the same emotion time series, creating dependence between group definition and the reported dynamic patterns. Independent validation (e.g., hold-out data or sensitivity to clustering parameters) is needed to rule out artifacts driving the inflexibility findings.

minor comments (2)

[Abstract] Abstract: Sample sizes are stated but no statistical details (p-values, effect sizes) accompany the group differences in outcomes or network influence.
[Methods] Notation: The free parameters of VISTA-SSM are not enumerated; a brief table or list would clarify reproducibility of the clustering step.

Simulated Author's Rebuttal

3 responses · 0 unresolved

We thank the referee for the detailed and constructive comments, which have helped strengthen the manuscript. We have revised the paper to incorporate performance metrics for the emotion model, expanded statistical reporting with baselines and confidence intervals, and added independent validation analyses to address potential circularity. Below we respond to each major comment.

read point-by-point responses

Referee: [Abstract/Methods] Abstract and Methods: No accuracy, F1, or inter-annotator metrics are reported for the transformer-based small language model on clinical therapy transcripts. This is load-bearing, as misclassification rates above ~20-30% for key emotions (sadness, fear) would propagate directly into the VISTA-SSM clusters, temporal networks, and the reported outcome associations for the n=3,813 non-response group.

Authors: We agree that explicit validation metrics for the emotion extraction model are necessary to evaluate robustness. In the revised manuscript we have added a dedicated subsection in Methods reporting accuracy, macro-F1, and inter-annotator agreement on a held-out sample of 500 clinical transcripts annotated by two licensed clinicians. The model achieves an overall accuracy of 0.81 and macro-F1 of 0.78; sadness and fear specifically reach F1 scores of 0.82 and 0.75. These values fall below the 20-30% error threshold the referee flags, supporting the stability of the downstream clusters and networks. revision: yes
Referee: [Results] Results: The manuscript provides no baseline comparisons, error bars, confidence intervals, or details on how symptom outcome measures were statistically linked to the emotion-derived clusters. Without these, the claims of increased deterioration odds and lower improvement likelihood in the non-response group cannot be evaluated for robustness.

Authors: We have substantially expanded the Results section. We now include (1) a baseline comparison using k-means clustering on baseline PHQ-9/GAD-7 scores alone, (2) error bars (standard errors) on all proportion estimates, and (3) 95% confidence intervals around the reported odds ratios and improvement probabilities. Statistical linkage is described via logistic regression models that predict binary outcome status (deterioration vs. no deterioration; clinically significant improvement vs. no improvement) from cluster membership while controlling for baseline severity, age, and session count. All associations remain statistically significant (p < 0.01) with the added controls. revision: yes
Referee: [Methods] Methods (VISTA-SSM): Clustering and network construction are performed on the same emotion time series, creating dependence between group definition and the reported dynamic patterns. Independent validation (e.g., hold-out data or sensitivity to clustering parameters) is needed to rule out artifacts driving the inflexibility findings.

Authors: We acknowledge the risk of circularity. In the revision we report two additional analyses performed on a 70/30 train/hold-out split of patients: (a) VISTA-SSM clustering was fit exclusively on the training set and temporal networks were then constructed and compared on the unseen hold-out set, reproducing the same pattern of elevated sadness/fear influence in the non-response group; (b) sensitivity checks varying the number of latent states (3–6) and the regularization parameter of the state-space model yield qualitatively identical network topologies. These results are now presented in a new subsection of Methods and an accompanying supplementary figure. revision: yes

Circularity Check

0 steps flagged

Minor dependence from shared emotion data in clustering and networks, but outcomes independent

full rationale

The derivation extracts emotions via transformer, applies VISTA-SSM to cluster dynamics and build temporal networks from the same linguistic data, then associates resulting groups with separate symptom outcome measures (deterioration odds, clinical improvement). This creates statistical dependence between clusters and networks but does not reduce the central claim to a self-definition or fitted input by construction. No equations, self-citations, or ansatzes are shown that force the result; the reported associations rely on external symptom data and thus retain independent content. Score reflects the reader's noted dependence without meeting load-bearing circularity thresholds.

Axiom & Free-Parameter Ledger

1 free parameters · 1 axioms · 0 invented entities

The central claim rests on the accuracy of pre-trained transformer emotion labeling and the validity of state-space assumptions for capturing temporal emotion dynamics; no independent evidence for either is provided in the abstract.

free parameters (1)

VISTA-SSM clustering parameters
Used to partition patients into improving and non-response subgroups from emotion time series

axioms (1)

domain assumption Small transformer models can reliably extract discrete emotion categories from conversational therapy text at turn level
Foundation for all downstream dynamics analysis

pith-pipeline@v0.9.0 · 5489 in / 1283 out tokens · 51546 ms · 2026-05-16T14:36:58.362190+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/RealityFromDistinction.lean reality_from_one_distinction unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

A transformer-based small language model extracted patients' emotions at the talk-turn level; a state-space model (VISTA-SSM) clustered subgroups based on emotion dynamics over time and produced temporal networks.

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

49 extracted references · 49 canonical work pages · 2 internal anchors

[1]

Nature Mental Health 1(12), 950–955 (2023)

Counts, N.Z., Bloom, D.E., Halfon, N.: Psychological distress as a systemic economic risk in the usa. Nature Mental Health 1(12), 950–955 (2023)

work page 2023
[2]

JAMA Network Open 8(10), 2540065–2540065 (2025)

Pullmann, M.D., Rouvere, J., Raue, P.J., Fillipo, I.R.G., Mosser, B.A., Heagerty, P.J., Fridling-Cook, N., Padmanabhan, A., Hull, T.D., Areán, P.A.: Message- based vs video-based psychotherapy for depression: A randomized clinical trial. JAMA Network Open 8(10), 2540065–2540065 (2025)

work page 2025
[3]

BMC psychiatry 20(1), 297 (2020)

Hull, T.D., Malgaroli, M., Connolly, P.S., Feuerstein, S., Simon, N.M.: Two-way messaging therapy for depression and anxiety: longitudinal response trajectories. BMC psychiatry 20(1), 297 (2020)

work page 2020
[4]

The Lancet Digital Health 7(3), 172–174 (2025)

Doherty, A., Bucci, S., Kenny, A., Kotov, R., Lipinska, G., Ospina-Pinillos, L., Schultebraucks, K.: Passive sensing at scale to transform understanding of poor mental health. The Lancet Digital Health 7(3), 172–174 (2025)

work page 2025
[5]

Translational Psychiatry 13(1), 309 (2023)

Malgaroli, M., Hull, T.D., Zech, J.M., Althoff, T.: Natural language processing for mental health interventions: a systematic review and research framework. Translational Psychiatry 13(1), 309 (2023)

work page 2023
[6]

Journal of Medical Internet Research 23(7), 28244 (2021)

Burkhardt, H.A., Alexopoulos, G.S., Pullmann, M.D., Hull, T.D., Areán, P.A., Cohen, T.: Behavioral activation and depression symptomatology: longitudinal assessment of linguistic indicators in text-based therapy sessions. Journal of Medical Internet Research 23(7), 28244 (2021)

work page 2021
[7]

Psychotherapy Research 31(3), 300–312 (2021)

Ewbank, M.P., Cummins, R., Tablan, V., Catarino, A., Buchholz, S., Blackwell, A.D.: Understanding the relationship between patient language and outcomes in internet-enabled cognitive behavioural therapy: A deep learning approach to automatic coding of session transcripts. Psychotherapy Research 31(3), 300–312 (2021)

work page 2021
[8]

In: Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology, pp

Shapira, N., Atzil-Slonim, D., Tuval-Mashiach, R., Shapira, O.: Measuring lin- guistic synchrony in psychotherapy. In: Proceedings of the Eighth Workshop on Computational Linguistics and Clinical Psychology, pp. 158–176 (2022)

work page 2022
[9]

Journal of Affective Disorders 352, 133–137 (2024)

Malgaroli, M., Hull, T.D., Calderon, A., Simon, N.M.: Linguistic markers of anxi- ety and depression in somatic symptom and related disorders: observational study of a digital intervention. Journal of Affective Disorders 352, 133–137 (2024)

work page 2024
[10]

Proceedings of the National Academy of Sciences 119(13), 2114737119 (2022)

Nook, E.C., Hull, T.D., Nock, M.K., Somerville, L.H.: Linguistic measures of psychological distance track symptom levels and treatment outcomes in a large set of psychotherapy transcripts. Proceedings of the National Academy of Sciences 119(13), 2114737119 (2022)

work page 2022
[11]

Psychotherapy Research 35(2), 174–189 (2025)

Eberhardt, S.T., Schaffrath, J., Moggia, D., Schwartz, B., Jaehde, M., Rubel, 17 J.A., Lutz, W.: Decoding emotions: Exploring the validity of sentiment analysis in psychotherapy. Psychotherapy Research 35(2), 174–189 (2025)

work page 2025
[12]

Clinical Psychology Review, 102641 (2025)

Luo, X., Levendosky, A.A.: A systematic review and integrative framework of psychotherapy microprocess: Linking the science of psychological interventions with the art of moment-to-moment practice. Clinical Psychology Review, 102641 (2025)

work page 2025
[13]

Journal of Consulting and Clinical Psychology (2024)

Paz, A., Rafaeli, E., Bar-Kalifa, E., Gilboa-Schechtman, E., Gannot, S., Narayanan, S.S., Atzil-Slonim, D.: Multimodal analysis of temporal affective variability within treatment for depression. Journal of Consulting and Clinical Psychology (2024)

work page 2024
[14]

Journal of Child and Adolescent Psychopharmacology 22(1), 37–47 (2012)

Forbes, E.E., Stepp, S.D., Dahl, R.E., Ryan, N.D., Whalen, D., Axelson, D.A., Silk, J.S.: Real-world affect and social context as predictors of treatment response in child and adolescent depression and anxiety: an ecological momentary assess- ment study. Journal of Child and Adolescent Psychopharmacology 22(1), 37–47 (2012)

work page 2012
[15]

Journal of Affective Disorders 206, 305–314 (2016)

Husen, K., Rafaeli, E., Rubel, J.A., Bar-Kalifa, E., Lutz, W.: Daily affect dynam- ics predict early response in cbt: Feasibility and predictive validity of ema for outpatient psychotherapy. Journal of Affective Disorders 206, 305–314 (2016)

work page 2016
[16]

Journal of Consulting and Clinical Psychology 92(8), 517 (2024)

Hehlmann, M.I., Moggia, D., Schwartz, B., Driver, C., Eberhardt, S., Lutz, W.: Outcome prediction in psychological therapy with continuous time dynamic mod- eling of affective states and emotion regulation. Journal of Consulting and Clinical Psychology 92(8), 517 (2024)

work page 2024
[17]

British Journal of Clinical Psychology 51(2), 206–222 (2012)

Wichers, M., Lothmann, C., Simons, C.J., Nicolson, N.A., Peeters, F.: The dynamic interplay between negative and positive emotions in daily life predicts response to treatment in depression: a momentary assessment study. British Journal of Clinical Psychology 51(2), 206–222 (2012)

work page 2012
[18]

Behaviour Research and Therapy 48(8), 754–760 (2010)

Peeters, F., Berkhof, J., Rottenberg, J., Nicolson, N.A.: Ambulatory emotional reactivity to negative daily life events predicts remission from major depressive disorder. Behaviour Research and Therapy 48(8), 754–760 (2010)

work page 2010
[19]

Perspectives on psychological science 8(6), 591–612 (2013)

Bonanno, G.A., Burton, C.L.: Regulatory flexibility: An individual differences perspective on coping and emotion regulation. Perspectives on psychological science 8(6), 591–612 (2013)

work page 2013
[20]

Motivation and Emotion 40(4), 602–624 (2016)

Coifman, K.G., Flynn, J.J., Pinto, L.A.: When context matters: Negative emo- tions predict psychological health and adjustment. Motivation and Emotion 40(4), 602–624 (2016)

work page 2016
[21]

Journal of abnormal 18 psychology 119(3), 479 (2010)

Coifman, K.G., Bonanno, G.A.: When distress does not become depression: emo- tion context sensitivity and adjustment to bereavement. Journal of abnormal 18 psychology 119(3), 479 (2010)

work page 2010
[22]

Perspectives on psychological science 6(3), 222–233 (2011)

Gruber, J., Mauss, I.B., Tamir, M.: A dark side of happiness? how, when, and why happiness is not always good. Perspectives on psychological science 6(3), 222–233 (2011)

work page 2011
[23]

Psychological science 15(7), 482–487 (2004)

Bonanno, G.A., Papa, A., Lalande, K., Westphal, M., Coifman, K.: The impor- tance of being flexible: The ability to both enhance and suppress emotional expression predicts long-term adjustment. Psychological science 15(7), 482–487 (2004)

work page 2004
[24]

Psychological science 24(12), 2505–2514 (2013)

Troy, A.S., Shallcross, A.J., Mauss, I.B.: A person-by-situation approach to emo- tion regulation: Cognitive reappraisal can either help or hurt, depending on the context. Psychological science 24(12), 2505–2514 (2013)

work page 2013
[25]

Nature Reviews Psychology 2(11), 663–675 (2023)

Bonanno, G.A., Chen, S., Galatzer-Levy, I.R.: Resilience to potential trauma and adversity through regulatory flexibility. Nature Reviews Psychology 2(11), 663–675 (2023)

work page 2023
[26]

European journal of Psychotraumatology 12(1), 1942642 (2021)

Bonanno, G.A.: The resilience paradox. European journal of Psychotraumatology 12(1), 1942642 (2021)

work page 2021
[27]

International Statistical Review (2024)

Lu, Z.: Clustering longitudinal data: A review of methods and software packages. International Statistical Review (2024)

work page 2024
[28]

Small Language Models are the Future of Agentic AI

Belcak, P., Heinrich, G., Diao, S., Fu, Y., Dong, X., Muralidharan, S., Lin, Y.C., Molchanov, P.: Small language models are the future of agentic ai. arXiv preprint arXiv:2506.02153 (2025)

work page internal anchor Pith review Pith/arXiv arXiv 2025
[29]

Psychological methods (2025)

Brindle, B., Hull, T.D., Malgaroli, M., Charon, N.: Vista-ssm: Varying and irreg- ular sampling time-series analysis via state space models. Psychological methods (2025)

work page 2025
[30]

Journal of general internal medicine 16(9), 606–613 (2001)

Kroenke, K., Spitzer, R.L., Williams, J.B.: The phq-9: validity of a brief depres- sion severity measure. Journal of general internal medicine 16(9), 606–613 (2001)

work page 2001
[31]

Annals of internal medicine 146(5), 317–325 (2007)

Kroenke, K., Spitzer, R.L., Williams, J.B., Monahan, P.O., Löwe, B.: Anxiety disorders in primary care: prevalence, impairment, comorbidity, and detection. Annals of internal medicine 146(5), 317–325 (2007)

work page 2007
[32]

https://huggingface.co/ j-hartmann/emotion-english-distilroberta-base/ (2022)

Hartmann, J.: Emotion English DistilRoBERTa-base. https://huggingface.co/ j-hartmann/emotion-english-distilroberta-base/ (2022)

work page 2022
[33]

RoBERTa: A Robustly Optimized BERT Pretraining Approach

Liu, Y., Ott, M., Goyal, N., Du, J., Joshi, M., Chen, D., Levy, O., Lewis, M., Zettlemoyer, L., Stoyanov, V.: Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692 (2019) 19

work page internal anchor Pith review Pith/arXiv arXiv 1907
[34]

Computers in Human Behavior 135, 107345 (2022)

Butt, S., Sharma, S., Sharma, R., Sidorov, G., Gelbukh, A.: What goes on inside rumour and non-rumour tweets and their reactions: A psycholinguistic analyses. Computers in Human Behavior 135, 107345 (2022)

work page 2022
[35]

ACM Transactions on Computing for Healthcare 6(2), 1–16 (2025)

Pathak, A., Bhattacharjee, S., Saha, T., Saha, S.: Do sentiment and emotion affect mental health? a multi-task classification framework for comprehensive understanding of mental health, emotion, and sentiment from motivational conversations. ACM Transactions on Computing for Healthcare 6(2), 1–16 (2025)

work page 2025
[36]

IEEE transactions on automatic control 9(4), 333–339 (1964)

Ho, Y.-C., Lee, R.: A bayesian approach to problems in stochastic estimation and control. IEEE transactions on automatic control 9(4), 333–339 (1964)

work page 1964
[37]

Pattern Recognition 138, 109375 (2023)

Umatani, R., Imai, T., Kawamoto, K., Kunimasa, S.: Time series clustering with an EM algorithm for mixtures of linear Gaussian state space models. Pattern Recognition 138, 109375 (2023)

work page 2023
[38]

The annals of mathematical statistics, 50–60 (1947)

Mann, H.B., Whitney, D.R.: On a test of whether one of two random variables is stochastically larger than the other. The annals of mathematical statistics, 50–60 (1947)

work page 1947
[39]

Pubbli- cazioni del R istituto superiore di scienze economiche e commericiali di firenze 8, 3–62 (1936)

Bonferroni, C.: Teoria statistica delle classi e calcolo delle probabilita. Pubbli- cazioni del R istituto superiore di scienze economiche e commericiali di firenze 8, 3–62 (1936)

work page 1936
[40]

Journal of Consulting and Clinical Psychology 59(1), 12–19 (1991)

Jacobson, N.S., Truax, P.: Clinical significance: a statistical approach to defining meaningful change in psychotherapy research. Journal of Consulting and Clinical Psychology 59(1), 12–19 (1991)

work page 1991
[41]

: Network analysis of multivariate data in psychological science

Borsboom, D., Deserno, M.K., Rhemtulla, M., Epskamp, S., Fried, E.I., McNally, R.J., Robinaugh, D.J., Perugini, M., Dalege, J., Costantini, G., et al. : Network analysis of multivariate data in psychological science. Nature Reviews Methods Primers 1(1), 58 (2021)

work page 2021
[42]

Springer, New York, NY (2003)

Ben-Israel, A., Greville, T.N.: Generalized Inverses: Theory and Applications. Springer, New York, NY (2003)

work page 2003
[43]

Journal of abnormal psychology 125(6), 747 (2016)

Robinaugh, D.J., Millner, A.J., McNally, R.J.: Identifying highly influential nodes in the complicated grief network. Journal of abnormal psychology 125(6), 747 (2016)

work page 2016
[44]

Clinical Psychology Review 85, 102000 (2021)

Malgaroli, M., Calderon, A., Bonanno, G.A.: Networks of major depressive disorder: A systematic review. Clinical Psychology Review 85, 102000 (2021)

work page 2021
[45]

Journal of Happiness studies 8(3), 371–392 (2007)

Coifman, K.G., Bonanno, G.A., Rafaeli, E.: Affect dynamics, bereavement and resilience to loss. Journal of Happiness studies 8(3), 371–392 (2007)

work page 2007
[46]

BMC psychiatry 19(1), 59 (2019)

Heininga, V.E., Dejonckheere, E., Houben, M., Obbels, J., Sienaert, P., Leroy, B., 20 Roy, J., Kuppens, P.: The dynamical signature of anhedonia in major depressive disorder: positive emotion dynamics, reactivity, and recovery. BMC psychiatry 19(1), 59 (2019)

work page 2019
[47]

Current Opinion in Psychology 17, 22–26 (2017)

Kuppens, P., Verduyn, P.: Emotion dynamics. Current Opinion in Psychology 17, 22–26 (2017)

work page 2017
[48]

Perspectives on Psychological Science 15(2), 444–468 (2020)

Lange, J., Dalege, J., Borsboom, D., Kleef, G.A., Fischer, A.H.: Toward an inte- grative psychometric model of emotions. Perspectives on Psychological Science 15(2), 444–468 (2020)

work page 2020
[49]

Proceedings of the National Academy of Sciences 114(10), 2016–2025 (2017) 21

LeDoux, J.E., Brown, R.: A higher-order theory of emotional consciousness. Proceedings of the National Academy of Sciences 114(10), 2016–2025 (2017) 21

work page 2016