Nikola Sekulovski
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Research Subjects
Nikola Sekulovski has a total of 68 co-authors. Here are the top co-authors by number of shared publications:
- Maarten Marsman
Department of Psychology, University of Amsterdam
18 shared publications - Sara Keetelaar
Department of Psychology, University of Amsterdam, Netherlands
10 shared publications - Eric-Jan Wagenmakers
Department of Psychology, University of Amsterdam, Amsterdam, Netherlands
7 shared publications Karoline Huth
7 shared publicationsDon van Den Bergh
5 shared publications- Denny Borsboom
Department of Psychological Methods, University of Amsterdam
4 shared publications
Latest Publications
- Open Access
- 22 October 2026
- Published by Center for Open Science in PsyArXiv
The Prevalence of Bayesian Inference, Preregistration, and Open Data in Psychology, 2004-2024
Psychology has undergone substantial methodological changes in the past two decades. We tracked the adoption of Bayesian inference and two open science practices across 28,738 articles from six ‘top-tier’ (15,634 articles) and six ‘mid-tier’ (13,104 articles) psychology journals in 2004–2024. In ‘top-tier’ journals, human (50%) and LLM-assisted (50%) full-text coding revealed that in 2024, 26% of articles reported a Bayesian analysis, 41% included a preregistration, and 73% provided open data. In ‘mid-tier’ journals, LLM coding showed corresponding rates of 16%, 18%, and 40%. From 2004 to 2012, these practices were virtually nonexistent. Bayesian logistic growth modeling suggests further increases, although their extent remains uncertain. Mixed-effects models became more prevalent, software reporting improved, with greater use of R and less reliance on SPSS, but estimation-focused (vs. hypothesis-testing-focused) reporting showed only modest gains. Despite heterogeneity across journals, psychology articles published in 2024 differ markedly from those published only a decade earlier.
- 22 October 2026
- Published by Center for Open Science in PsyArXiv
Open AccessThe Prevalence of Bayesian Inference, Preregistration, and Open Data in Psychology, 2004-2024
Psychology has undergone substantial methodological changes in the past two decades. We tracked the adoption of Bayesian inference and two open science practices across 28,738 articles from six ‘top-tier’ (15,634 articles) and six ‘mid-tier’ (13,104 articles) psychology journals in 2004–2024. In ‘top-tier’ journals, human (50%) and LLM-assisted (50%) full-text coding revealed that in 2024, 26% of articles reported a Bayesian analysis, 41% included a preregistration, and 73% provided open data. In ‘mid-tier’ journals, LLM coding showed corresponding rates of 16%, 18%, and 40%. From 2004 to 2012, these practices were virtually nonexistent. Bayesian logistic growth modeling suggests further increases, although their extent remains uncertain. Mixed-effects models became more prevalent, software reporting improved, with greater use of R and less reliance on SPSS, but estimation-focused (vs. hypothesis-testing-focused) reporting showed only modest gains. Despite heterogeneity across journals, psychology articles published in 2024 differ markedly from those published only a decade earlier.
- Open Access
- 21 October 2026
- Published by SAGE Publications in Advances in Methods and Practices in Psychological Science
- Vol. 9 (4)
Analysis Blinding as a Complement to Preregistration: A Tutorial for Network Analysis
Network analysis has become a popular computational-modeling approach in psychological research and is now increasingly used to address confirmatory research questions. However, preregistration of such analyses remains rare and even when implemented may leave room for researcher degrees of freedom, particularly when data-dependent decisions are difficult to fully anticipate. In this tutorial, we demonstrate how researchers can safeguard against bias by using analysis blinding as a complementary step to preregistration. The method involves temporarily altering the data to remove the key effect of interest while preserving all other aspects. Analysis blinding makes it possible to explore important features of the data, for instance, spotting outliers or adjusting the computational model, without introducing bias. In this tutorial, we present three common research questions in network analysis and show how they can be addressed using analysis blinding to safeguard their confirmatory nature. We use data from previously published studies employing network analysis and offer a practical, step-by-step guide along with R code that researchers can adapt to their own data and research questions.
- 21 October 2026
- Published by SAGE Publications in Advances in Methods and Practices in Psychological Science
- Vol. 9 (4)
Open AccessAnalysis Blinding as a Complement to Preregistration: A Tutorial for Network Analysis
Network analysis has become a popular computational-modeling approach in psychological research and is now increasingly used to address confirmatory research questions. However, preregistration of such analyses remains rare and even when implemented may leave room for researcher degrees of freedom, particularly when data-dependent decisions are difficult to fully anticipate. In this tutorial, we demonstrate how researchers can safeguard against bias by using analysis blinding as a complementary step to preregistration. The method involves temporarily altering the data to remove the key effect of interest while preserving all other aspects. Analysis blinding makes it possible to explore important features of the data, for instance, spotting outliers or adjusting the computational model, without introducing bias. In this tutorial, we present three common research questions in network analysis and show how they can be addressed using analysis blinding to safeguard their confirmatory nature. We use data from previously published studies employing network analysis and offer a practical, step-by-step guide along with R code that researchers can adapt to their own data and research questions.
- Open Access
- 20 October 2026
- Published by Center for Open Science in PsyArXiv
Analysis and Reporting Guidelines for Bayesian Graphical Modeling of Cross-Sectional Data: A Tutorial in JASP
Bayesian graphical modeling (BGM) is an emerging approach to psychological network analysis that offers proper uncertainty quantification and allows for incorporating prior information. However, resources that integrate the full analytic workflow within user-friendly software and provide guidance for transparent reporting remain scarce. In this tutorial, we address this gap by offering an integrated analysis and reporting guide for BGM. We first briefly outline the theoretical foundations of the framework. We then introduce the updated Bayesian network analysis module in the open-source software JASP and demonstrate three core analyses using an empirical dataset: edge selection and parameter estimation, clustering analysis, and comparing networks across groups. For each analysis, we provide minimal template paragraphs for the Methods and Results sections. We conclude with recommendations for prior robustness analyses and open research practices.
- 20 October 2026
- Published by Center for Open Science in PsyArXiv
Open AccessAnalysis and Reporting Guidelines for Bayesian Graphical Modeling of Cross-Sectional Data: A Tutorial in JASP
Bayesian graphical modeling (BGM) is an emerging approach to psychological network analysis that offers proper uncertainty quantification and allows for incorporating prior information. However, resources that integrate the full analytic workflow within user-friendly software and provide guidance for transparent reporting remain scarce. In this tutorial, we address this gap by offering an integrated analysis and reporting guide for BGM. We first briefly outline the theoretical foundations of the framework. We then introduce the updated Bayesian network analysis module in the open-source software JASP and demonstrate three core analyses using an empirical dataset: edge selection and parameter estimation, clustering analysis, and comparing networks across groups. For each analysis, we provide minimal template paragraphs for the Methods and Results sections. We conclude with recommendations for prior robustness analyses and open research practices.
- Open Access
- 26 June 2026
- Published by Center for Open Science in PsyArXiv
A Stochastic Block Prior for Clustering in Graphical Models
Graphical models facilitate the representation of psychological variables as complex systems of interacting variables structured as a network. However, their current statistical estimation procedures overlook the assumption of clustering, which refers to the grouping of subsets of variables that are more densely connected within the network, despite this assumption playing a central role in many psychological theories. We address this gap by proposing the use of the Stochastic Block Model (SBM) as a prior distribution on the network structure of a graphical model for binary and ordinal data. The SBM assumes that variables belong to latent clusters, where the probability of an edge depends on the cluster membership of the nodes. Embedding this prior in a Bayesian graphical modeling framework allows researchers to formally incorporate theoretical expectations about clustering, test hypotheses about the number of clusters, and estimate cluster assignments from cross-sectional data. We demonstrate the benefits of this approach in a simulation study and reanalyze 30 openly available empirical datasets to test for clustering. This work highlights how the Bayesian framework can embed theoretical assumptions into network models via priors and introduces a new tool for latent cluster inference in psychological network analysis.
- 26 June 2026
- Published by Center for Open Science in PsyArXiv
Open AccessA Stochastic Block Prior for Clustering in Graphical Models
Graphical models facilitate the representation of psychological variables as complex systems of interacting variables structured as a network. However, their current statistical estimation procedures overlook the assumption of clustering, which refers to the grouping of subsets of variables that are more densely connected within the network, despite this assumption playing a central role in many psychological theories. We address this gap by proposing the use of the Stochastic Block Model (SBM) as a prior distribution on the network structure of a graphical model for binary and ordinal data. The SBM assumes that variables belong to latent clusters, where the probability of an edge depends on the cluster membership of the nodes. Embedding this prior in a Bayesian graphical modeling framework allows researchers to formally incorporate theoretical expectations about clustering, test hypotheses about the number of clusters, and estimate cluster assignments from cross-sectional data. We demonstrate the benefits of this approach in a simulation study and reanalyze 30 openly available empirical datasets to test for clustering. This work highlights how the Bayesian framework can embed theoretical assumptions into network models via priors and introduces a new tool for latent cluster inference in psychological network analysis.
- 25 June 2026
- Published by American Psychological Association (APA) in Psychological Methods
A stochastic block prior for clustering in graphical models.
Graphical models facilitate the representation of psychological variables as complex systems of interacting variables structured as a network. However, their current statistical estimation procedures overlook the assumption of clustering, which refers to the grouping of subsets of variables that are more densely connected within the network, despite this assumption playing a central role in many psychological theories. We address this gap by proposing the use of the stochastic block model as a prior distribution on the network structure of a graphical model for binary and ordinal data. The stochastic block model assumes that variables belong to latent clusters, where the probability of an edge depends on the cluster membership of the nodes. Embedding this model in a Bayesian graphical modeling framework allows researchers to formally incorporate theoretical expectations about clustering, test hypotheses about the number of clusters, and estimate cluster membership of the nodes from cross-sectional data. We illustrate the performance of the method in a simulation study and apply it to a reanalysis of 30 empirical datasets. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
- 25 June 2026
- Published by American Psychological Association (APA) in Psychological Methods
A stochastic block prior for clustering in graphical models.
Graphical models facilitate the representation of psychological variables as complex systems of interacting variables structured as a network. However, their current statistical estimation procedures overlook the assumption of clustering, which refers to the grouping of subsets of variables that are more densely connected within the network, despite this assumption playing a central role in many psychological theories. We address this gap by proposing the use of the stochastic block model as a prior distribution on the network structure of a graphical model for binary and ordinal data. The stochastic block model assumes that variables belong to latent clusters, where the probability of an edge depends on the cluster membership of the nodes. Embedding this model in a Bayesian graphical modeling framework allows researchers to formally incorporate theoretical expectations about clustering, test hypotheses about the number of clusters, and estimate cluster membership of the nodes from cross-sectional data. We illustrate the performance of the method in a simulation study and apply it to a reanalysis of 30 empirical datasets. (PsycInfo Database Record (c) 2026 APA, all rights reserved).
- Open Access
- 30 March 2026
- Published by Center for Open Science in PsyArXiv
LLM-Based Prior Elicitation for Bayesian Graphical Modeling
In the Bayesian Graphical Modeling framework, priors on network structureencode theoretical assumptions and uncertainty about the topology of psychologicalconstructs under study. For instance, the Bernoulli prior specifiesthe probability of each pairwise interaction, the Beta–Bernoulli prior governsexpected network density, and the Stochastic Block prior models clustering.In practice, however, specifying informed hyperparameters is challenging:theoretical guidance is limited, and default choices can be overly simplisticor restrictive. To address this, we introduce an LLM-based prior elicitationframework in which a large language model provides inclusion judgments foreach variable pair. These judgments are converted into edge-specific priorprobabilities for the Bernoulli prior and used to derive hyperparameters forthe Beta–Bernoulli and Stochastic Block priors. To make the approach accessible,we provide an R package, bgmElicit, with a Shiny app implementingthe methodology. We illustrate the framework in two examples. First, avalidation on a subset of a PTSD network from a meta-analysis comparesOpenAI GPT models across several conditions. Second, an empirical analysisof 17 PTSD symptoms shows that elicited priors can modestly strengthenevidence regarding edge presence and absence. Taken together, this work isa proof of concept, complementary to expert judgment and prior sensitivitychecks.
- 30 March 2026
- Published by Center for Open Science in PsyArXiv
Open AccessLLM-Based Prior Elicitation for Bayesian Graphical Modeling
In the Bayesian Graphical Modeling framework, priors on network structureencode theoretical assumptions and uncertainty about the topology of psychologicalconstructs under study. For instance, the Bernoulli prior specifiesthe probability of each pairwise interaction, the Beta–Bernoulli prior governsexpected network density, and the Stochastic Block prior models clustering.In practice, however, specifying informed hyperparameters is challenging:theoretical guidance is limited, and default choices can be overly simplisticor restrictive. To address this, we introduce an LLM-based prior elicitationframework in which a large language model provides inclusion judgments foreach variable pair. These judgments are converted into edge-specific priorprobabilities for the Bernoulli prior and used to derive hyperparameters forthe Beta–Bernoulli and Stochastic Block priors. To make the approach accessible,we provide an R package, bgmElicit, with a Shiny app implementingthe methodology. We illustrate the framework in two examples. First, avalidation on a subset of a PTSD network from a meta-analysis comparesOpenAI GPT models across several conditions. Second, an empirical analysisof 17 PTSD symptoms shows that elicited priors can modestly strengthenevidence regarding edge presence and absence. Taken together, this work isa proof of concept, complementary to expert judgment and prior sensitivitychecks.
- Open Access
- 30 March 2026
- Published by Wiley in British Journal of Mathematical and Statistical Psychology
LLM‐based prior elicitation for Bayesian graphical modeling
In the Bayesian graphical modeling framework, priors on network structure encode theoretical assumptions and uncertainty about the topology of psychological constructs under study. For instance, the Bernoulli prior specifies the probability of each pairwise interaction, the Beta–Bernoulli prior governs expected network density, and the Stochastic Block prior models clustering. In practice, however, specifying informed hyperparameters is challenging: theoretical guidance is limited, and default choices can be overly simplistic or restrictive. To address this, we introduce an LLM‐based prior elicitation framework in which a large language model provides inclusion judgments for each variable pair. These judgments are converted into edge‐specific prior probabilities for the Bernoulli prior and used to derive hyperparameters for the Beta–Bernoulli and Stochastic Block priors. To make the approach accessible, we provide an R package, bgmElicit , with a Shiny app implementing the methodology. We illustrate the framework in two examples. First, a validation on a subset of a PTSD network from a meta‐analysis compares OpenAI GPT models across several conditions. Second, an empirical analysis of 17 PTSD symptoms shows that elicited priors can modestly strengthen evidence regarding edge presence and absence. Taken together, this work is a proof of concept, complementary to expert judgment and prior sensitivity checks.
- 30 March 2026
- Published by Wiley in British Journal of Mathematical and Statistical Psychology
Open AccessLLM‐based prior elicitation for Bayesian graphical modeling
In the Bayesian graphical modeling framework, priors on network structure encode theoretical assumptions and uncertainty about the topology of psychological constructs under study. For instance, the Bernoulli prior specifies the probability of each pairwise interaction, the Beta–Bernoulli prior governs expected network density, and the Stochastic Block prior models clustering. In practice, however, specifying informed hyperparameters is challenging: theoretical guidance is limited, and default choices can be overly simplistic or restrictive. To address this, we introduce an LLM‐based prior elicitation framework in which a large language model provides inclusion judgments for each variable pair. These judgments are converted into edge‐specific prior probabilities for the Bernoulli prior and used to derive hyperparameters for the Beta–Bernoulli and Stochastic Block priors. To make the approach accessible, we provide an R package, bgmElicit , with a Shiny app implementing the methodology. We illustrate the framework in two examples. First, a validation on a subset of a PTSD network from a meta‐analysis compares OpenAI GPT models across several conditions. Second, an empirical analysis of 17 PTSD symptoms shows that elicited priors can modestly strengthen evidence regarding edge presence and absence. Taken together, this work is a proof of concept, complementary to expert judgment and prior sensitivity checks.
- Open Access
- 14 November 2025
- Published by Center for Open Science in PsyArXiv
The Impact of Dichotomization on Network Recovery
Graphical models have become an important method for studying the network structure of multivariate psychological data. Accurate recovery of the underlying network structure is paramount and requires that the models are appropriate for the data at hand. Traditionally, Gaussian graphical models for continuous data and Ising models for binary data have dominated the literature. However, psychological research often relies on ordinal data from Likert scale items, creating a model-data mismatch. This paper examines the effect of dichotomizing ordinal variables on network recovery, as opposed to analyzing the data at its original level of measurement, using a Bayesian analysis of the ordinal Markov random field model. This model is implemented in the R package bgms. Our analysis shows that dichotomization results in a loss of information, which affects the accuracy of network recovery. This is particularly true when considering the interplay between the dichotomization cutoffs used and the distribution of the ordinal categories. In addition, we show that there is a difference in accuracy when using dichotomized data depending on whether edges are included or excluded in the true network, highlighting the effectiveness of using the ordinal model in recovering conditional independence relationships. These findings underscore the importance of using models that deal directly with ordinal data to ensure more reliable and valid inferred network structures in psychological research.
- 14 November 2025
- Published by Center for Open Science in PsyArXiv
Open AccessThe Impact of Dichotomization on Network Recovery
Graphical models have become an important method for studying the network structure of multivariate psychological data. Accurate recovery of the underlying network structure is paramount and requires that the models are appropriate for the data at hand. Traditionally, Gaussian graphical models for continuous data and Ising models for binary data have dominated the literature. However, psychological research often relies on ordinal data from Likert scale items, creating a model-data mismatch. This paper examines the effect of dichotomizing ordinal variables on network recovery, as opposed to analyzing the data at its original level of measurement, using a Bayesian analysis of the ordinal Markov random field model. This model is implemented in the R package bgms. Our analysis shows that dichotomization results in a loss of information, which affects the accuracy of network recovery. This is particularly true when considering the interplay between the dichotomization cutoffs used and the distribution of the ordinal categories. In addition, we show that there is a difference in accuracy when using dichotomized data depending on whether edges are included or excluded in the true network, highlighting the effectiveness of using the ordinal model in recovering conditional independence relationships. These findings underscore the importance of using models that deal directly with ordinal data to ensure more reliable and valid inferred network structures in psychological research.
- Open Access
- 11 November 2025
- Published by Center for Open Science in PsyArXiv
Safeguarding Against Bias Without Preregistration: A Tutorial on Analysis Blinding for Network Analysis
Network analysis has become a popular computational modeling approach in psychological research and is now increasingly used to address confirmatory research questions. However, preregistration of such analyses remains rare and given the complexity of the data and methods, is often perceived as limiting or even infeasible. In this tutorial, we demonstrate how researchers can safeguard against bias without relying on preregistration by using analysis blinding. The method involves temporarily altering the data to remove the key effect of interest while preserving all other aspects. Analysis blinding makes it possible to explore important features of the data, for instance, spotting outliers or adjusting the computational model, without introducing bias. This tutorial presents three common research questions in network analysis and shows how they can be addressed using analysis blinding to safeguard their confirmatory nature. It uses data from previously published studies employing network analysis, and offers a practical, step-by-step guide along with R code that researchers can adapt to their own data and research questions.
- 11 November 2025
- Published by Center for Open Science in PsyArXiv
Open AccessSafeguarding Against Bias Without Preregistration: A Tutorial on Analysis Blinding for Network Analysis
Network analysis has become a popular computational modeling approach in psychological research and is now increasingly used to address confirmatory research questions. However, preregistration of such analyses remains rare and given the complexity of the data and methods, is often perceived as limiting or even infeasible. In this tutorial, we demonstrate how researchers can safeguard against bias without relying on preregistration by using analysis blinding. The method involves temporarily altering the data to remove the key effect of interest while preserving all other aspects. Analysis blinding makes it possible to explore important features of the data, for instance, spotting outliers or adjusting the computational model, without introducing bias. This tutorial presents three common research questions in network analysis and shows how they can be addressed using analysis blinding to safeguard their confirmatory nature. It uses data from previously published studies employing network analysis, and offers a practical, step-by-step guide along with R code that researchers can adapt to their own data and research questions.
- Open Access
- 11 November 2025
- Published by Springer Nature in Behavior Research Methods
- Vol. 57 (12) , 1-11
The impact of dichotomization on network recovery
Graphical models have become an important method for studying the network structure of multivariate psychological data. Accurate recovery of the underlying network structure is paramount and requires that the models are appropriate for the data at hand. Traditionally, Gaussian graphical models for continuous data and Ising models for binary data have dominated the literature. However, psychological research often relies on ordinal data from Likert scale items, creating a model-data mismatch. This paper examines the effect of dichotomizing ordinal variables on network recovery, as opposed to analyzing the data at its original level of measurement, using a Bayesian analysis of the ordinal Markov random field model. This model is implemented in the R package . Our analysis shows that dichotomization results in a loss of information, which affects the accuracy of network recovery. This is particularly true when considering the interplay between the dichotomization cutoffs used and the distribution of the ordinal categories. In addition, we demonstrate a difference in accuracy when using dichotomized data, depending on whether edges are included or excluded in the true network, which highlights the effectiveness of the ordinal model in recovering conditional independence relationships. These findings underscore the importance of using models that deal directly with ordinal data to ensure more reliable and valid inferred network structures in psychological research.
- 11 November 2025
- Published by Springer Nature in Behavior Research Methods
- Vol. 57 (12) , 1-11
Open AccessThe impact of dichotomization on network recovery
Graphical models have become an important method for studying the network structure of multivariate psychological data. Accurate recovery of the underlying network structure is paramount and requires that the models are appropriate for the data at hand. Traditionally, Gaussian graphical models for continuous data and Ising models for binary data have dominated the literature. However, psychological research often relies on ordinal data from Likert scale items, creating a model-data mismatch. This paper examines the effect of dichotomizing ordinal variables on network recovery, as opposed to analyzing the data at its original level of measurement, using a Bayesian analysis of the ordinal Markov random field model. This model is implemented in the R package . Our analysis shows that dichotomization results in a loss of information, which affects the accuracy of network recovery. This is particularly true when considering the interplay between the dichotomization cutoffs used and the distribution of the ordinal categories. In addition, we demonstrate a difference in accuracy when using dichotomized data, depending on whether edges are included or excluded in the true network, which highlights the effectiveness of the ordinal model in recovering conditional independence relationships. These findings underscore the importance of using models that deal directly with ordinal data to ensure more reliable and valid inferred network structures in psychological research.