99 Idiographic personality networks 5 1. Introduction Personality is traditionally conceptualized in terms of traits that are relatively stable across situations and over time (Allport, 1937; cf. Mischel & Shoda, 1995). Observations of within-person temporal patterns, however, show far more variability than stability over time. In fact, few if any people respond to stimuli completely equally across different and seemingly similar situations over time (Shoda et al., 1994). To account for these idiosyncrasies, personality processes ought to be modeled for each individual separately. Recently this became possible through the introduction of statistically estimated idiographic personality networks (Beck & Jackson, 2020; Costantini et al., 2019; Lazarus et al., 2020; Springstein & English, 2023). Idiographic network models are estimated from intensive longitudinal within-person data, such as ecological momentary assessments (EMA), which are visualized as a person-specific network of statistical associations between different personality components and their interdependencies. Yet, the degree to which idiographic network structures can inform personality theory and research remains unclear. This paper will employ conventional idiographic personality network analysis, with the aim to demonstrate that the stability of these networks is not just theoretically unlikely but also empirically dubious, which may pose a problem for studying individual differences. Stability, in this paper, exclusively implies time-invariance within-persons, not across persons. We first introduce how idiographic personality networks are employed in studies of individual differences, before we elaborate on potential disconnects between personality theory and the assumptions of this new type of model. Idiographic personality networks are intuitively understandable graphs in which various personality components are represented as nodes, and pairwise interdependencies between them are represented as edges; connections between these nodes (Cramer et al., 2012). Personality components are selfreported behaviors, cognitions and emotions derived from personality surveys, which are taken to be constitutive of traits (Cramer et al., 2012). It is worth noting that idiographic network analysis differs from psychometric network analysis of personality trait surveys (e.g., Borsboom et al., 2021; Chen et al., 2023; Christensen et al., 2020) in the sense that it applies to one individual and does not aim to map out the population-level personality structure. That is, in idiographic network analysis the interdependency between nodes is estimated with pairwise partial correlations between time-series of one individual. Statistically estimating the edges is done on two timescales: contemporaneous and lag-1 associations (Epskamp & Fried, 2018), which respectively are 1) covariances of each personality component at time-point t with other personality components at the same time-point t and 2) covariances between
RkJQdWJsaXNoZXIy MTk4NDMw