iiinova — where your digital self keeps evolving

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 09/10/2026

Congratulations, Amanda! Very well-deserved honours!

010

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 08/10/2026

Happy you got a good laugh, Yannick! :-) Hope all is well in France!

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

Haha, sorry about this! :') Recently finished a longer stretch of teaching and other major commitments, and finally had some decent, sustained time to push through several things that had been sitting at 98% done and in the pipeline for far too long!

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

If subgroups depend on when we happen to measure, they can't reliably guide precision care cts-gimme clusters people by their underlying dynamics, far less tied to the measurement schedule, & provides a key step towards reliable dynamic profiling in psy sci & mental health 🌐 osf.io/preprints/ps...

osf.io

OSF

040

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

We tested the method 73 patients with depression, assessed 5x/day for 28 days ▶️ Shared dynamics: anhedonia, lethargy & restlessness predicting depressed mood Subgroup features: ▶️ Profile 1: mood–restlessness cycle ▶️ Profile 2: vicious cycle linking mood & lethargy, & lethargy --> irritability

130

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

In the simulation study, cts-gimme outperformed its discrete-time counterpart, when subgroup differences were subtle (realistic scenario in psyc) ▶️ ARI .74–.96 vs .04–.53 ▶️ 80% vs 26% of pair classifications correct across all sampling rates w/larger differences+sparse sampling, S-GIMME did better

130

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

Without pre-specifying subgroups, it identifies key levels of shared and individual dynamics: ▶️ Dynamics shared across the full sample (everyone in the population) ▶️ Recurring dynamics shared within subgroups ▶️ Dynamics unique to each individual

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

We introduce a subgrouping extension to continuous-time Group Iterative Multiple Model Estimation (cts-gimme) that clusters individuals by their dynamics in continuous time. cts-gimme models how variables change continuously, independent of when they were measured, and handles irregular intervals.

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

A key problem is that most subgrouping methods work in discrete time But clustering based on discrete-time methods depends on the assessment schedule, revealing different subgroups of dynamics due to measurement Who gets grouped together can then reflect study design & not the underlying processes

130

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

A key question is whether individual differences in dynamics also show recurring similarities across individuals (Ebrahimi et al, 2024). After a long-term collaboration w/@jonathanpark.bsky.social, Mills & H. Riese, we present a method for reliably identifying subgroups with shared dynamics.

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 07/10/2026

⚡ Out now: Psychological and health sciences increasingly study how phenomena dynamically influence and reinforce one another over time: for example, to understand the mechanisms maintaining mental disorders. Can we identify reliable clusters of these dynamics? A thread 🧵 osf.io/preprints/ps...

25320

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

In conclusion: How a loss is remembered (specific memory characteristics & their interactions) help keep grief alive. Treating grief may partly require changing how that loss is held in memory. Led by the brilliant F. Lechner-Meichsner & Kirsten Smith, w/Anke Ehlers 🌐: doi.org/10.1159/0005...

doi.org

When Memories Keep Grief Alive: A Dynamic Model of Memory Characteristics and Prolonged Grief

Abstract. Introduction: Prolonged grief disorder (PGD) is a condition marked by sustained separation distress. Although cognitive models suggest that memory processes are critical in PGD, the specific...

130

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

Cognitive-behavioural theories place memory processes at centre of prolonged grief. These findings provide support for that account & add specificity, highlighting: ▶️ How specific memory characteristics reinforce one-another over time ▶️ How different memory features map onto specific PGD symptoms

121

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

Other features mapped onto specific symptoms: ▶️ Negative memory predominance --> yearning (the only predictor of this core symptom) ▶️ Visceral reactions --> loneliness (grief felt in the body can be hard to explain to others, and easier to feel alone in)

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

Re-experiencing and emotional numbness also predicted increases in each other over time: indicative of a vicious cycle. Numbing is a form of avoidance. It stops the loss memory from being processed, so the memory stays raw and immediate, and that rawness invites more shutting down.

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

Re-experiencing was the most influential memory characteristic. Memories that felt vivid and 'happening now' predicted: ▶️ Later difficulty accepting the death ▶️ More memory triggers, which predicted difficulty moving forward

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

Main results: Memory characteristics predicted later grief symptoms far more than the reverse. 6 lagged effects ran from memory to symptoms; only 1 ran from symptoms to memory. The most influential nodes were memory features; the most impacted were symptoms.

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

275 bereaved adults were assessed longitudinally up to 18 months after the loss. Using dynamic network modelling (panel graphicalVAR), we modelled how within-person fluctuations in memory characteristics and grief symptoms dynamically interacted with one-another over time.

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

Memories of a loss differ in how they are experienced. Using the Oxford Grief-Memory Characteristics Scale, we examined 6 features, e.g.: ▶️ Re-experiencing: vivid, feels like it is happening now ▶️ Triggers: set off by many cues ▶️ Visceral: felt in the body as pain or fatigue

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

Around 1 in 20 bereaved people develop prolonged grief disorder (PGD): persistent, impairing yearning for or preoccupation with the deceased. Cognitive theories have proposed that memory processes keep grief alive. We wanted to test which memory processes predict which specific symptoms over time?

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 30/09/2026

📜 Out in Psychotherapy & Psychosomatics Most bereaved people adapt over time. Why do some stay stuck in grief? We tested whether how loss is remembered shapes grief symptoms over the first 18 months of bereavement w/Franziska Lechner-Meichsner, Anke Ehlers & Kirsten Smith doi.org/10.1159/0005...

1168

Reposted by Omid V. Ebrahimi

Jonathan J. Park @jonathanpark.bsky.social · 25/09/2026

New preprint! People with different underlying processes may appear similar simply because of when we measure them; which we refer to as “illusory homogeneity” We introduce cts-gimme to identify group-, subgroup-, and person-specific dynamics in continuous-time: osf.io/preprints/ps...

osf.io

OSF

1113

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

Centrality is not a standalone measure of “importance” Its meaning – and whether it is consequential – depends on the topology it is embedded in, and on how change propagates through that system 👥 w/ @jonathanpark.bsky.social, @mijke.bsky.social, @dennyborsboom.bsky.social 🌐 osf.io/preprints/ps...

osf.io

OSF

020

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

Centrality also depended on context Strength usually produced the clearest separation, but no metric was universally best: ▶️ At very low thresholds, almost any node triggered a cascade ▶️ At very high, almost none did So centrality matters when network topology lets it translate into wider effects

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

💡 Under moderate activation thresholds, starting a cascade at the most versus least central node mattered most in Barabási–Albert networks. Why? The “most central” node was often an exceptional hub.

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

🛠️ To test this, we used a weighted cascading-failure model: We activated one node, then let activation spread when enough of its neighbours were active. We then compared how far cascades spread when starting from the most vs least central node across the different network topologies.

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

3 network structures with the same number of nodes and edges were simulated: ▶️ Erdős–Rényi: random edge placement ▶️ Watts–Strogatz: local clustering + shortcuts ▶️ Barabási–Albert: preferential attachment + hubs We then asked: when does activating the most central node actually change what happens?

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

Centrality is widely used to identify “important” nodes in psychological networks. But being #1 does not always convey consequential information (e.g.: cf. spread of activation). Could this depend on the underlying structure of the network & mechanisms of spread?

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

Centrality was key in epi networks. The brilliant @jonathanpark.bsky.social vowed to one day ask when centrality has consequential meaning in psych networks. 5 yrs later, I’ve had the pleasure [& pain] of being Jonathan’s [Batman] Robin, in collab w/@mijke.bsky.social & @dennyborsboom.bsky.social

120

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

@jonathanpark.bsky.social & I met online in Prof. Sy-Miin Chow’s Dynamical Systems class in 2021, during a COVID lockdown. While bonding over doing PhDs in a global crisis, we watched infectious disease networks play out in real time: Some nodes could matter enormously for how infection spread.

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 24/09/2026

Tired of the centrality debate? Of seeing one application where centrality is influential & others where it's more arbitrary? We have the article for you: osf.io/preprints/ps... Turns out centrality can indeed matter; but when it does depends on underlying network structure & mechanism of spread 🧵

1216

Reposted by Omid V. Ebrahimi

Adela Isvoranu @adelaisvoranu.bsky.social · 23/09/2026

I've been noticing that “cross-sectional” has become shorthand for “methodologically weak” in (network) research, and often data being cross-sectional is treated as a methodological objection in itself. I took some time to put my thoughts down in writing. Pre-print: osf.io/preprints/ps... 🧵⬇️

osf.io

OSF

3277

Reposted by Omid V. Ebrahimi

Aleksi Hamina @aleksihamina.bsky.social · 21/09/2026

Very creative and impressive work!

141

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 22/09/2026

Thanks for your kind words, Aleksi!

010

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

Implications: Effective treatment may work partly by decoupling symptoms: weakening the dynamic dependencies in which one problem amplifies another This suggests a complementary way to study treatment mechanisms: not only symptom reduction, but changes in system dynamics 🌐 doi.org/10.1192/bjp....

doi.org

Dopamine antagonists and dynamic symptom networks in schizophrenia | The British Journal of Psychiatry | Cambridge Core

Dopamine antagonists and dynamic symptom networks in schizophrenia

030

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

Led by the brilliant Rob McCutcheon (@rob-mccutcheon.bsky.social) w/collaborators Elisabeth Kenny & Bodyl Brand This has been a rewarding one to collaborate on personally & a hypothesis I’ve wanted to test for a long time We recently also wrote about this in another context: osf.io/preprints/ps...

osf.io

OSF

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

This advances our understanding of change during treatment as: ▶️ Clinical improvement may involve decoupling symptoms, not only reducing them. ▶️ Strongly propagating symptoms may be targets for adjunctive care. ▶️ Network metrics can be tested as additional treatment response markers.

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

This matters because weakened cross-symptom coupling means that a change (e.g. activation or aggravation) in one symptom is less strongly linked to subsequent changes (presence or increase) in others. Treatment may therefore reduce the system’s tendency to propagate activation across symptoms.

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

Notably, this was driven by weakened cross-symptom connections: 10.71 (treatment) vs 17.89 (placebo), p = .004. But symptom self-persistence did not differ: 8.21 vs 8.64, p = .352. ▶️ Treatment weakened how symptoms spread to one another, not how persistent each symptom was on its own.

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

Main results: Symptom connectivity was significantly (p < .001) lower in individuals taking paliperidone (18.92) than the placebo group (26.52): ▶️ This suggests that symptoms become less tightly coupled over time during treatment: they reinforce each other less.

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

We pooled 8 double-blind RCTs of paliperidone (antipsychotic) vs placebo: ▶️ 2508 people w/schizophrenia or schizoaffective disorder ▶️ 20 PANSS symptoms ▶️ Assessed at baseline & weeks 2, 4 & 6 Using panel graphicalVAR, we modelled how each symptom predicted later changes in itself & other symptoms

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

This leads to a different treatment question. Beyond asking whether symptoms become less severe, can treatment change how symptoms impact and reinforce each other over time? In other words: can treatment weaken the links through which symptoms propagate over time?

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

Symptoms of psychosis don't occur in isolation: Hallucination can worsen anxiety, while persecutory delusions can lead to withdrawal. Network theory (@dennyborsboom.bsky.social) proposes such interactions help maintain disorders. Effective treatment should therefore also weaken these interactions

110

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 21/09/2026

📜 Out in British Journal of Psychiatry What if antipsychotic treatment works not only by reducing symptoms, but also by weakening how strongly symptoms drive one another? w/McCutcheon, Kenny & Brand We tested this using 8 randomized trials of paliperidone in 2508 adults 🧵 doi.org/10.1192/bjp....

doi.org

Dopamine antagonists and dynamic symptom networks in schizophrenia | The British Journal of Psychiatry | Cambridge Core

Dopamine antagonists and dynamic symptom networks in schizophrenia

1124

Reposted by Omid V. Ebrahimi

Pam Davis-Kean, PhD @umpamdk.bsky.social · 15/09/2026

My newest Presidential Column for @psychscience.bsky.social addresses the issue of psychologists' dominant use of convenience samples for their research. Can our science stop being WEIRD if we don't consider how we sample? www.psychologicalscience.org/publications...

psychologicalscience.org

Is It Too Convenient to Stay WEIRD?

APS President Pamela Davis-Kean outlines her plenary topics for the 2027 APS Annual Convention in Seattle, Washington.

01912

Reposted by Omid V. Ebrahimi

Ludvig Daae Bjørndal @ludvigdb.bsky.social · 19/09/2026

I am thrilled to share this work - with the brilliant @michelnivard.bsky.social and @omidvebrahimi.bsky.social - in which we outline a framework to more closely align psychiatric genetics and aetiological research with the dynamics and complexities of mental disorders 🎯

093

Reposted by Omid V. Ebrahimi

Jonathan J. Park @jonathanpark.bsky.social · 18/09/2026

New preprint! Psychometric networks are often summarized with centrality indices that are sometimes useful and other times not. We show that the "most central" node's influence depends heavily on the configuration of edges in the network; its degree distribution. osf.io/preprints/ps...

osf.io

OSF

2197

Omid V. Ebrahimi @omidvebrahimi.bsky.social · 18/09/2026

Very interesting and insightful study for longitudinal researchers. Many thanks!

120

Reposted by Omid V. Ebrahimi

Michel Nivard @michelnivard.bsky.social · 17/09/2026

I worked with @omidvebrahimi.bsky.social and @ludvigdb.bsky.social on bringing the systems perspective of psych, “too” psychiatric genetics. it has been a journey to figure out where the fields are aligned, where in place but hidden by terminology & how to translate reseach Q from field to field!

1172