关闭

当前位置:首页 > 微信群推荐 » 正文

Estimating psychopathological networks: Be careful what you

Network models, in which psychopathological disorders are conceptualized as a complex interplay of psychological and biological components, have become increasingly popular in the recent psychopathological literature (Borsboom, et. al., 2011). These network models often contain significant numbers of unknown parameters, yet the sample sizes available in psychological research are limited. As such, general assumptions about the true network are introduced to reduce the number of free parameters. Incorporating these assumptions, however, means that the resulting network will lead to reflect the particular structure assumed by the estimation method—a crucial and often ignored aspect of psychopathological networks. For example, observing a sparse structure and simultaneously assuming a sparse structure does not imply that the true model is, in fact, sparse. To illustrate this point, we discuss recent literature and show the effect of the assumption of sparsity in three simulation studies.

Citation: Epskamp S, Kruis J, Marsman M (2017) Estimating psychopathological networks: Be careful what you wish for. PLoS ONE 12(6): e0179891. https://doi.org/10.1371/journal.pone.0179891

Editor: Daniele Marinazzo, Ghent University, BELGIUM

Received: April 26, 2016; Accepted: June 2, 2017; Published: June 23, 2017

Copyright: © 2017 Epskamp et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Data Availability: All relevant data are within the paper and Supporting Information files.

Funding: This work was supported by NWO (Netherlands Organisation for Scientific Research) “research talent” Grant Number 406-11-06 (). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Competing interests: The authors have declared that no competing interests exist.

Introduction

Recent psychological literature has focused on a network approach to model many different psychological phenomena []. Such networks can be high-dimensional structures (i.e., the number of unknown parameters is much larger than the available data), which are hard to estimate without making general assumptions about the underlying true model structure. If the true model is assumed to be sparse, thus containing a small number of connections relative to the number of nodes, a methodology can be applied that potentially returns a sparse network structure. In other words, assuming a sparse network structure results in estimating a sparse network structure, which means that certain conclusions cannot be drawn from observing such a structure. In this paper, we argue that care should be taken in interpreting the obtained network structure because the estimation procedure may pollute the results. We will illustrate this by showing examples of networks obtained when sparse networks are estimated even when the true network structure is dense.

Network psychometrics

The network approach has been particularly promising in the field of psychopathology. Within this framework, symptoms (e.g., insomnia, fatigue, and concentration problems) are no longer treated as interchangeable indicators of some latent mental disorder (e.g., depression). Instead, symptoms play an active causal role. For example, insomnia leads to fatigue, fatigue leads to concentration problems, and so forth []. Psychopathological disorders, then, are not interpreted as the common cause of observed symptoms but rather as emergent behaviors that result from a complex interplay of psychological and biological components. To grasp such a complex structure, a network model can be used in which variables such as symptoms or moods are represented by nodes. Nodes are connected by edges that indicate associations between nodes. This line of research has led to intuitive new insights about various psychopathological concepts such as comorbidity [, ], the impact of life events [, ], and sudden life transitions (e.g., sudden onset of a depressive episode; [, ]). For an overview of network modeling applied to psychopathology, we refer the reader to a recent review of Fried et al. [].

The growing popularity of the network perspective on psychological phenomena has culminated in the emergence of a new branch of psychology dedicated to the estimation of network structures on psychological data—network psychometrics []. This field focuses on tackling the problem of estimating network structures involving large numbers of parameters in high-dimensional models. When cross-sectional data are analyzed, the most popular models that are used are the Gaussian Graphical Model (GGM; []) for continuous data and the Ising model [] for binary data. Both the GGM and the Ising model fall under a general class of models called Markov Random Fields. These models represent variables as nodes which are connected by edges but only if the variables are conditionally independent. The strength of an edge (i.e., its absolute deviance from zero) demonstrates the strength of the association between two variables after conditioning on all other variables in the network; this is also termed concentration []. In the GGM, edges directly correspond to partial correlation coefficients. The Ising model does not allow for such standardization, but edge weights can be similarly interpreted. A more detailed introduction of network models is beyond the scope of this paper, but we recommend [] and [] for further reading on the subject.

上一篇:手机里竟然有这么多传感器!终于都搞懂了
下一篇:中文(简体)
说点什么吧
  • 全部评论(0
    还没有评论,快来抢沙发吧!

猜你喜欢


怎么查看手机的ip地址

怎么查看手机的ip地址,很多人不知道手机也有i地址。i全称IteretProtocol。意思是网络之间的协议。所以只要联网的设备...

二维码