103 Idiographic personality networks 5 the average changes halfway from 0 in the first half to 6 in the second half. This non-stationarity invalidates a summary statistic like the total average (3), which does justice to neither the first half, second half nor the whole process. Similar to the average in this univariate example, average network estimates are invalid when the weak stationarity assumption is violated. The consequence is that they could gravely misrepresent the actual dynamic process. Weak stationarity is thus necessary to interpret an idiographic personality network as representative of the underlying personality processes. However, examining (non–)stationarity in networks is not straightforward, because it is possible that a linear combination of multiple non-stationary time-series results in stable average relations between them (cf. cointegration; Hamilton, 1994; Ryan et al., 2023). Currently there is not enough empirical research to verify whether the (weak) stationarity assumption holds in idiographic personality networks. Findings from the few available research (i.e., within-person network variability found in Beck & Jackson, 2021) points in the opposite direction: non-stationarity in the data-generating processes. Examining (non)stationarity further is imperative because we rely on idiographic networks to provide a valid description of someone’s personality system, which then forms the basis of inferences about differences in personality structures between people. The current study will explore how variable personality network structures are between individuals and within individuals over time. This illustrative study is based on a sample of adolescents and young individuals with a mild intellectual disability or borderline intellectual functioning who participated in a 60-day daily diary study. We first explore the degree of homogeneity of idiographic networks between individuals. Second, we explore homogeneity within subgroups of individuals who, based on traditional personality screening, share a personality profile. Third, we explore network homogeneity within-persons. More specifically, for each individual we assess 1) how variable or stable the networks are over time and 2) whether there is stability in its variability (i.e., stationarity). Based on these three research questions we discuss implications for network modeling and theory building.
RkJQdWJsaXNoZXIy MTk4NDMw