Registration for the 2021 Psychological Networks Amsterdam Winter School is now open! See here for more details and registration.
Last month, we hosted the first online version of our summer school on network analysis. All materials except solutions are now available on the OSF and – for a limited time (until October 31) – you can now get access via Eventbrite to 50+ video lectures for free.
Special thanks to everyone who helped make this possible: Sacha Epskamp, Adela Isvoranu, Ria Hoekstra, Denny Borsboom, Riet van Bork, Julian Burger, Jonas Haslbeck, Lourens Waldorp, Eiko Fried, Alessandra Mansueto, Karoline Huth, Jill de Ron, Adam Finnemann, Gaby Lunansky, Jolanda Van der Ree-Kossakowski, Pia Tio, and Lourens Waldorp.
On May 7 2020 we published the paper ‘The polarization within and across individuals: the hierarchical Ising opinion model’ in the Journal of Complex networks.
Polarization of opinions involves psychological processes as well as group dynamics. However, the interaction between the within individual dynamics of attitude formation and across person polarization is rarely studied. By modelling individual attitudes as Ising networks of attitude elements, and approximating this behaviour by the cusp singularity, we developed a fundamentally new model of social dynamics.
In this hierarchical model, agents behave either discretely or continuously depending on their attention to the issue. At the individual level the model reproduces the mere thought effect and resistance to persuasion. At the social level the model implies polarization and the persuasion paradox. We propose a new intervention for escaping polarization in bounded confidence models of opinion dynamics.
Han van der Maas, Jonas Dalege, and Lourens Waldorp
Psychological Methods, UvA
There are several exciting job vacancies in Amsterdam on topics very closely related to our research: one assistant professor position at the Psychological Methods Group, and several PhD positions hosted through the newly founded Center for Urban Mental Health. Note that PhD positions in the Netherlands are full-time 4-year paid jobs including the same benefits as any other job (e.g., pension buildup). Starting a PhD position in the Netherlands requires a Master’s degree. In this blog post, I will highlight some of the most relevant positions. Please consider applying if you are eligible, or forwarding these vacancies to eligible candidates you may know!
Assistant Professor of Psychological Methods
The first position I would like to point out is an assistant professor position at the Psychological Methods Group. The Psychological Methods group of the University of Amsterdam is one of the largest and most successful research and education centers in the field of psychological methods. In the past decade, several strong lines of research originated from this group, such as network modeling of psychological phenomena in the Psychosystems and the Psychonetrics lab groups, Bayesian statistical tools implemented in the open-source statistics program JASP, and adaptive testing implemented in the spin-off company Oefenweb. We have recently launched the increasingly popular Behavioral Data Science master program, which will be the focus of this assistant professorship as well.
Please note that this position does not come with a tenure-track (these do not exist at the Psychological Methods Group), but does come with an outlook on a permanent position!
PhD positions at the Center for Urban Mental Health
The University of Amsterdam has recently approved the foundation of the first-ever Center for Urban Mental Health, which will form an interdisciplinary approach to tackling common mental health problems from a complexity point of view. This center will at first be housed at the renowned Institute for Advanced Studies, right in the city center of Amsterdam. The Center for Urban Mental Health will be launched with several interdisciplinary PhD projects, all aiming to start early 2020. Below, I will list the most relevant positions that are currently open for applications!
Computational modeling of psychological and social dynamics in urban mental health conditions: the case of addictive substance use.
The first PhD vacancy is a PhD project between the Psychological Methods Group in the Department of Psychology, the Department of Computational Science in the Informatics Institute, and the Institute for Advanced Studies. This project will be supervised by me (Sacha Epskamp) and Michael Lees, and will be hosted at the Psychological Methods group and the Institute for Advanced Studies. The aim of this project is to form computational models (e.g., differential equations, agent based models, Ising models), that combine psychological dynamics with social dynamics. This means that we wish to form models that can simulate intraindividual dynamics of multiple people that also interact with each other in complex ways. We are specifically looking for a candidate with a background in using such models as well as a strong affinity with psychological research.
Please note that the closing date is already this Sunday (November 24). However, if you are interested in doing a PhD on the topic of computational modeling of psychological dynamics, feel free to contact me (Sacha Epskamp) as more positions related to this topic may become available.
Network Analysis and Urban Mental Health Interventions
The second PhD vacancy is a vacancy between the Department of Psychology and the Department of Communication Science. The project will be supervised by me (Sacha Epskamp), alongside Reinout Wiers, Julia van Weert, and Barbara Schouten. This project will start out as a methodological project, investigating ways of estimating network models from self-report time-series data (e.g., experience sampling method; ESM), after which the project will move towards a more applied clinical focus in which the candidate is expected to gather and analyze ESM data, followed by collaborating with clinicians to derive and evaluate intervention plans. We are specifically looking for a candidate that has both a strong focus on clinical research as well as good data analytic skills and an affinity or background in methodology.
Network theory of addiction and depression
The third PhD vacancy is a vacancy between the Psychological Methods Group in the Department of Psychology, the Department of Psychiatry at the Amsterdam UMC, location AMC, and the Institute for Advanced Studies. This project will be supervised by Maarten Marsman, Judy Luigjes, and Ruth van Holst. This project will build upon the view that depression and addictions form a complex system of mutually interacting problems and aims to formalize a network model of the two disorders in an urban context. The project will involve analyzing existing cross-sectional, longitudinal and clinical data, taken from an urban population. Where necessary, new techniques and models will be developed for the analysis of these data. The candidate should have a strong affinity to (clinical) psychological research and formal statistical modelling (e.g., Bayesian hierarchical modelling, or mathematical statistics), and proficiency in programming in at least R.
Even more positions!
If this was not enough, more positions will open up shortly! At the Center for Urban Mental Health! Currently there is a listing on the complexity of fear, depression and addiction among adolescents (in Dutch), and a listing on aging and mental health. And a few more positions are expected to open up. If you are interested in these, I encourage you to check the Urban Mental Health website and/or follow the Center for Urban Mental Health on Twitter. In addition, even more job opportunities may open up in the coming months, depending on various sources of funding. While I cannot go into detail on those, I highly recommend joining our very active Facebook group on Psychological Dynamics, in which both job opportunities on topics related to our research are routinely posted.
This blog post was written by Sacha Epskamp
I would like to share the following three updates:
1. In January we hosted the annual Psychological Networks Amsterdam Winter School. On popular request, we have now made many of the materials publicly available on the Open Science Framework: https://osf.io/he3wj.
2. I teach two courses on Structural Equation Modeling at the University of Amsterdam, which are closely related and often touch on network modeling as well. Many materials (including video summaries) of the latest course are now available at http://sachaepskamp.com/SEM2019
3. We have extended the deadline for our event in Singapore on complexities of adverse behavior and mental health (http://complexsystems.nl/ccs2019/) to June 16! Please share this with anyone you think might be interested!
According to the network theory of mental disorders, mental disorders are developed and sustained through direct interactions between symptoms . From this conceptualization it follows that treatment of mental disorders should involve changing the network of interrelated symptoms. While over the past years many studies have investigated the network structure of psychopathologies , the effect of psychological treatment on the network of interrelated symptoms has rarely been assessed. Moreover, network analysis techniques provide a unique opportunity to investigate treatment effects at a more detailed symptom level. In this new study we aspired to adopt and extent the network approach to investigate specific and sequential treatment effects—a technique we labelled Network Intervention Analysis (NIA) . NIA involves estimating a network of interrelated symptoms and including an additional treatment indicator variable that encodes the treatment condition that a participants belongs to. Including such a variable into the network allows to see which symptoms are conditionally dependent on this treatment variable. Since the treatment can influence the symptoms but not vice versa, this dependency indicates the symptoms that are directly affected by treatment.
In our illustration of NIA we investigated the effect of cognitive behavioral therapy for insomnia (CBTI) on co-occurring insomnia and depression symptoms. More traditional analyses had showed that after completion, the treatment had relieved both insomnia and depression symptoms . It remained unclear, however, whether the effect of treatment on the depression symptoms occurred via improving the sleep problems, or whether CBTI influenced the depression symptoms directly. Using NIA across 10 measurement weeks (2 prior to treatment, 5 during treatment, and 3 after treatment) we could identify that CBTI predominantly affected the sleep problems, indicating that the improvements in depression likely occur via CBTI-induced improvements in sleep.
Figure. Network structure before, during, and after treatment. The networks include the Insomnia Severity Index and Patient Health Questionnaire items (circles) and treatment (square). The size of the node is proportional to the difference in symptom severity between the treatment and control group, where smaller node sizes represent greater differences in favor of the treatment group. All ten networks corresponding to each of the measurement weeks are shown in the paper. An animated version can be found online.
This paper is a first illustration of how NIA can be used to investigate sequential and symptom-specific treatment induced changes over time. We hope to further develop NIA into a more sophisticated technique to investigate treatment effects over time—to ultimately better understand treatment mechanisms and reveal clues to their optimization.
 Borsboom, D. (2017). A network theory of mental disorders. World Psychiatry, 16, 5-13.
 Fried, E.I., Van Borkulo, C.D., Cramer, A.O.J., Boschloo, L., Schoevers, R.A., Borsboom, D. (2017). Mental disorders as networks of problems: A review of recent insights. Social Psychiatry and Psychiatric Epidemiology, doi.org/10.1007/s00127-016-1319-z.
 Blanken, T.F.*, Van der Zweerde, T.*, Van Straten, A., Van Someren, E.J.W., Borsboom, D., Lancee, J. (2019). Introducing Network Intervention Analysis to investigate sequential, symptom-specific treatment effects: A demonstration in co-occurring insomnia and depression. Psychotherapy and Psychosomatics, doi.org/10.1159/000495045.
 Van der Zweerde T, Van Straten A, Effting M, Kyle SD, Lancee J. (2018). Does online insomnia treatment reduce depressive symptoms? A randomized controlled trial in individuals with both insomnia and depressive symptoms. Psychological Medicine, 49, 501-599.
*shared first authors
Busy times at the Psychosystems group! In 2018, Max Hinne joined us as a postdoc, working on a position shared between Psychosystems and Eric-Jan Wagenmakers’ Bayesian research group (the people behind the fantastic JASP program, which now incorporates a network analysis module designed by Don van den Berg). Max studies ways to integrate information on network structure and network dynamics by utilizing Bayesian approaches. Also in 2018, NWO-Veni laureate Maarten Marsman, who works on network models in the context of educational measurement, was awarded a research fellowship at the Institute for Advanced Studies in Amsterdam. In addition, Ria Hoekstra and Julian Burger successfully applied for a Research Talent grant of the Netherlands Organization for Scientific Research and are starting their Ph. D. projects this month. Ria will study methods to address heterogeneity in network structures, with Denny Borsboom acting as promotor. Julian, whose primary location is at Groningen University with Robert Schoevers acting as promotor, will develop ways to translate dynamical systems theory and network modeling into tools that are useful in clinical practice. Sacha Epskamp, who (directly after winning the 2018 Psychometric Society Dissertation Award for his thesis on network psychometrics) secured an NWO-Veni project on network modeling which also happens to start this month, will be involved in the supervision of the new Ph. D. projects, acting as co-promotor and daily supervisor, while Maarten Marsman will also be involved in Julian’s project. Incidentally, Sacha was not the only person receiving praise for his Ph. D. thesis this year, as Claudia van Borkulo finished second in the Van Swinderen Prize 2018 for her dissertation on symptom network models in depression research. Finally, Jonas Dalege is transitioning into a postdoc position this month; he will work on Denny Borsboom’s ERC consolidator grant in a joint position with the Social Psychology department at the University of Amsterdam.
On Thursday 27th we will host the satellite session “Understanding Psychological Phenomena as Complex Systems” at the Conference on Complex Systems 2018, focusing on advancing the theoretical foundations of the network approach to psychopathology. Abstracts and slides are available on the satellite website!
This blog post was written by Sacha Epskamp.
In the past year, we have published two tutorial papers on the bootnet package for R, which aims to provide an encompassing framework for estimating network structures and checking their accuracy and stability. The first paper, published in Behavior Research Methods, focuses on how to use the bootstrapping methods to assess the accuracy and stability of network structures. The second, published in Psychological Methods, focuses on the most commonly used network model, the Gaussian graphical model (GGM; a network of partial correlations), and discusses further utility of the bootnet package.
With more than a year of development, version 1.1 of the bootnet package marks the most ambitious series of updates to the package to date. The version is now ready for testing on Github, and will soon be released on CRAN. This version includes or fleshes out a total of 7 new default sets, including one default set aimed at time-series analysis, and offers new functionality to the default sets already included in the package. In addition, the package now fully supports directed networks as well as methods resulting in multiple network models, and supports the expected influence centrality measure. Furthermore, the plotting functions in bootnet have been updated and now show more information. Finally, the package includes a new simulation function,
replicationSimulator, and greatly improves the functionality of the
netSimulator function. With these extensions, bootnet now aims to provide an exhaustive simulation suite for network models in many conditions. This blog post is intended to introduce this new functionality and to encourage testing of the functionality before the package is submitted to CRAN.
The package can be installed using:
Updates to network estimation
Bootnet now contains the following default sets:
|“EBICglasso”||GGM||Regularized GGM estimation using glasso and EBIC Model selection||qgraph, glasso|
|“pcor”||GGM||Unregularized GGM estimation, possibly thresholded using significance testing or FDR||qgraph|
|“IsingFit”||Ising||Regularized Ising models using nodewise logistic regressions and EBIC model selection||IsingFit, glmnet|
|“IsingSampler”||Ising||Unregularized Ising moddel estimation||IsingSampler|
|“huge”||GGM||Regularized GGM estimation using glasso and EBIC Model selection||huge|
|“adalasso”||GGM||Regularized GGM estimation using adaptive LASSO and cross-validation||parcor|
|“mgm”||Mixed graphical model||Regularized mixed graphical model estimation using EBIC or cross-validation||mgm||Revamped|
|“relimp”||Relative importance networks||Relative importance networks, possible using another default for network structure||relaimpo||Yes|
|“cor”||Correlation networks||Correlation networks, possibly thresholded using significance testing or FDR||qgraph||Yes|
|“TMFG”||Correlation networks & GGM||Triangulated Maximally Filtered Graph||NetworkToolbox||Yes|
|“LoGo”||GGM||Local/Global Sparse Inverse Covariance Matrix||NetworkToolbox||Yes|
|“ggmModSelect”||GGM||Unregularized (E) BIC model selection using glasso and stepwise estimation||qgraph, glasso||Yes|
|“graphicalVAR”||Graphical VAR||Regularized estimation of temporal and contemporaneous networks||graphicalVAR||Yes|
As can be seen, several of these are newly added. For example, we can now estimate the two recently added more conservative GGM estimation methods added to qgraph, which I described in an earlier blog post. Taking my favorite dataset as example:
The optimal unregularized network that minimizes BIC (note that tuning defaults to 0 in this case) can be obtained as follows:
net_modSelect <- estimateNetwork(bfi[,1:25], default = "ggmModSelect", stepwise = FALSE, corMethod = "cor")
It is generally recommended to use the stepwise improvement of edges with
stepwise = TRUE, but I set it to
FALSE here to make this tutorial easier to follow. Likewise, the thresholded regularized GGM can be obtained using the new
threshold argument for the EBICglasso default set:
net_thresh <- estimateNetwork(bfi[,1:25], tuning = 0, # EBICglasso sets tuning to 0.5 by default default = "EBICglasso", threshold = TRUE, corMethod = "cor")
As typical, we can plot the network:
Layout <- qgraph::averageLayout(net_modSelect, net_thresh) layout(t(1:2)) plot(net_modSelect, layout = Layout, title = "ggmModSelect") plot(net_thresh, layout = Layout, title = "Thresholded EBICglasso")
Principal direction and expected influence
One robust finding in psychological literature is that all variables tend to correlate positively after arbitrary rescaling of variables – the positive manifold. Likewise, it is common to expect parameters focusing on conditional associations (e.g., partial correlations) to also be positive after rescaling variables. Negative edges in such a network could indicate (a) violations of the latent variable model and (b) the presence of common cause effects in the data. To this end, bootnet now includes the ‘principalDirection’ argument in many default sets, which takes the first eigenvector of the correlation matrix and multiplies all variables by their corresponding sign. For example:
net_modSelect_rescale <- estimateNetwork(bfi[,1:25], default = "ggmModSelect", stepwise = FALSE, principalDirection = TRUE) net_thresh_rescale <- estimateNetwork(bfi[,1:25], tuning = 0, default = "EBICglasso", threshold = TRUE, principalDirection = TRUE) layout(t(1:2)) plot(net_modSelect_rescale, layout = Layout, title = "ggmModSelect") plot(net_thresh_rescale, layout = Layout, title = "Thresholded EBICglasso")
This makes the edges of an unexpected sign much more profound, at the cost of interpretability (as now the rescaling of some variables has to be taken into account).
One potentially useful centrality index that can be used on graphs mostly showing only positive relationships (and all variables recoded in the same direction) is the newly proposed expected influence measure. Expected influence computes node strength without taking the absolute value of edge-weights. This centrality measure can now be obtained:
qgraph::centralityPlot( list( ggmModSelect = net_modSelect_rescale, EBICGlasso_thresh = net_thresh_rescale ), include = "ExpectedInfluence" )
To bootstrap expected influence, it has to be requested from
boots <- bootnet(net_thresh_rescale, statistics = "ExpectedInfluence", nBoots = 1000, nCores = 8, type = "case") library("ggplot2") plot(boots, statistics = "ExpectedInfluence") + theme(legend.position = "none")
In addition to expected influence, thanks to the contribution of Alex Christensen, randomized shortest paths betweenness centrality (RSPBC) and hybrid centrality are now also supported, which can be called using
statistics = c("rspbc", "hybrid").
Relative importance networks
Relative importance networks can be seen as a re-parameterization of the GGM in which directed edges are used to quantify the (relative) contribution in predictability of each variable on other variables. This can be useful mainly for two reasons: (1) the relative importance measures are very stable, and (2) centrality indices based on relative importance networks have more natural interpretations. For example, in-strength of non-normalized relative importance networks equals \(R^2\). Bootnet has been updated to support directed networks, which allows for support for estimating relative importance networks. The estimation may be slow with over 20 variables though. Using only the first 10 variables of the BFI dataset we can compute the relative importance network as follows:
net_relimp <- estimateNetwork(bfi[,1:10], default = "relimp", normalize = FALSE)
As the relative importance network can be seen as a re-parameterization of the GGM, it makes sense to first estimate a GGM structure and subsequently impose that structure on the relative importance network estimation. This can be done using the structureDefault argument:
net_relimp2 <- estimateNetwork(bfi[,1:10], default = "relimp", normalize = FALSE, structureDefault = "ggmModSelect", stepwise = FALSE # Sent to structureDefault function )
Layout <- qgraph::averageLayout(net_relimp, net_relimp2) layout(t(1:2)) plot(net_relimp, layout = Layout, title = "Saturated") plot(net_relimp2, layout = Layout, title = "Non-saturated")
The difference is hardly visible because edges that are removed in the GGM structure estimation are not likely to contribute a lot of predictive power.
The LASSO regularized estimation of graphical vector auto-regression (VAR) models, as implemented in the graphicalVAR package, is now supported in bootnet! For example, we can use the data and codes supplied in the supplementary materials of our recent publication in Clinical Psychological Science to obtain the detrended data object
Data. Now we can run the graphical VAR model (I use
nLambda = 8 here to speed up computation, but higher values are recommended and are used in the paper):
# Variables to include: Vars <- c("relaxed","sad","nervous","concentration","tired","rumination", "bodily.discomfort") # Estimate model: gvar <- estimateNetwork( Data, default = "graphicalVAR", vars = Vars, tuning = 0, dayvar = "date", nLambda = 8 )
We can now plot both networks:
Layout <- qgraph::averageLayout(gvar$graph$temporal, gvar$graph$contemporaneous) layout(t(1:2)) plot(gvar, graph = "temporal", layout = Layout, title = "Temporal") plot(gvar, graph = "contemporaneous", layout = Layout, title = "Contemporaneous")
The bootstrap can be performed as usual:
gvar_boot <- bootnet(gvar, nBoots = 100, nCores = 8)
To plot the results, we need to make use of the
plot(gvar_boot, graph = "contemporaneous", plot = "interval")
Updates to bootstrapping methods
Splitting edge accuracy and model inclusion
Some of the new default sets (
LoGo) do not rely on regularization techniques to pull estimates to zero. Rather, they first select a set of edges to include, then estimate a parameter value only for the included edges. The default plotting method of edge accuracy, which has been updated to include the means of the bootstraps, will then not accurately reflect the range of parameter values. Making use of arguments
plot = "interval" and
split0 = TRUE, we can plot quantile intervals only for the times the parameter was not set to zero, in addition to a box indicating how often the parameter was set to zero:
# Agreeableness items only to speed things up: net_modSelect_A <- estimateNetwork(bfi[,1:5], default = "ggmModSelect", stepwise = TRUE, corMethod = "cor") # Bootstrap: boot_modSelect <- bootnet(net_modSelect_A, nBoots = 100, nCores = 8) # Plot results: plot(boot_modSelect, plot = "interval", split0 = TRUE)
This shows that the edge A1 (Am indifferent to the feelings of others) – A5 (Make people feel at ease) was always removed from the network, contradicting the factor model. The transparency of the intervals shows also how often an edge was included. For example, the edge A1 – A4 was almost never included, but when it was included it was estimated to be negative.
Accuracy of directed networks
Accuracy plots of directed networks are now supported:
net_relimp_A <- estimateNetwork(bfi[,1:5], default = "relimp", normalize = FALSE) boot_relimp_A <- bootnet(net_relimp_A, nBoots = 100, nCores = 8) plot(boot_relimp_A, order = "sample")
netSimulator function and accompanying plot method have been greatly expanded. The most basic use of the function is to simulate the performance of an estimation method given some network structure:
Sim1 <- netSimulator( input = net_modSelect, dataGenerator = ggmGenerator(), nCases = c(100,250,500,1000), nCores = 8, nReps = 100, default = "ggmModSelect", stepwise = FALSE) plot(Sim1)
Instead of keeping the network structure fixed, we could also use a function as
input argument to generate a different structure every time. The updated
genGGM function allows for many such structures (thanks to Mark Brandt!). In addition, we can supply multiple arguments to any estimation argument used to test multiple conditions. For example, perhaps we are interested in investigating if the stepwise model improvement is really needed in 10-node random networks with \(25\%\) sparsity:
Sim2 <- netSimulator( input = function()bootnet::genGGM(10, p = 0.25, graph = "random"), dataGenerator = ggmGenerator(), nCases = c(100,250,500,1000), nCores = 8, nReps = 100, default = "ggmModSelect", stepwise = c(FALSE, TRUE)) plot(Sim2, color = "stepwise")
which shows slightly better specificity at higher sample sizes. We might wish to repeat this simulation study for 50% sparse networks:
Sim3 <- netSimulator( input = function()bootnet::genGGM(10, p = 0.5, graph = "random"), dataGenerator = ggmGenerator(), nCases = c(100,250,500,1000), nCores = 8, nReps = 100, default = "ggmModSelect", stepwise = c(FALSE, TRUE)) plot(Sim3, color = "stepwise")
The results object is simply a data frame, meaning we can combine our results easily and use the plot method very flexibly:
Sim2$sparsity <- "sparsity: 0.75" Sim3$sparsity <- "sparsity: 0.50" Sim23 <- rbind(Sim2,Sim3) plot(Sim23, color = "stepwise", yfacet = "sparsity")
Investigating the recovery of centrality (correlation with true centrality) is also possible:
plot(Sim23, color = "stepwise", yfacet = "sparsity", yvar = c("strength", "closeness", "betweenness", "ExpectedInfluence"))
Likewise, we can also change the x-axis variable:
Sim23$sparsity2 <- gsub("sparsity: ", "", Sim23$sparsity) plot(Sim23, color = "stepwise", yfacet = "nCases", xvar = "sparsity2", xlab = "Sparsity")
This allows for setting up powerful simulation studies with minimal effort. For example, inspired by the work of Williams and Rast, we could compare such an un-regularized method with a regularized method and significance thresholding:
Sim2b <- netSimulator( input = function()bootnet::genGGM(10, p = 0.25, graph = "random"), dataGenerator = ggmGenerator(), nCases = c(100,250,500,1000), nCores = 8, nReps = 100, default = "EBICglasso", threshold = c(FALSE, TRUE)) Sim2c <- netSimulator( input = function()bootnet::genGGM(10, p = 0.25, graph = "random"), dataGenerator = ggmGenerator(), nCases = c(100,250,500,1000), nCores = 8, nReps = 100, default = "pcor", threshold = "sig", alpha = c(0.01, 0.05)) # add a variable: Sim2$estimator <- paste0("ggmModSelect", ifelse(Sim2$stepwise,"+step","-step")) Sim2b$estimator <- paste0("EBICglasso", ifelse(Sim2b$threshold,"+threshold","-threshold")) Sim2b$threshold <- NULL Sim2c$estimator <- paste0("pcor (a = ",Sim2c$alpha,")") # Combine: Sim2full <- dplyr::bind_rows(Sim2, Sim2b, Sim2c) # Plot: plot(Sim2full, color = "estimator")
In response to a number of studies aiming to investigate the replicability of network models, I have implemented a function derived from
netSimulator that instead simulates two datasets from a network model, treating the second as a replication dataset. This allows researchers to investigate what to expect when aiming to replicate effect. The input and usage is virtually identical to that of
SimRep <- replicationSimulator( input = function()bootnet::genGGM(10, p = 0.25, graph = "random"), dataGenerator = ggmGenerator(), nCases = c(100,250,500,1000), nCores = 8, nReps = 100, default = "ggmModSelect", stepwise = c(FALSE, TRUE)) plot(SimRep, color = "stepwise")
As always, I highly welcome bug reports and code suggestions on Github. I would also welcome any volunteers willing to help work in this project. This work can include adding new default sets, but also overhauling the help pages or other work. Please contact me if you are interested!