Scholarlink

Purpose <p>While psychological change processes are increasingly assumed to be “non-ergodic”, prompting a shift toward idiographic approaches, the assumption of ergodicity is often accepted a priori rather than tested empirically. This study investigates the validity of this assumption in the context of symptom dynamics during cognitive-behavioral therapy (CBT), focusing on the equivalence of network structures, i.e., structural ergodicity.</p> Methods <p>Using a sample of weekly transdiagnostic symptom assessments from <i>N</i> = 230 patients who received an average of 39.5 (<i>SD</i> = 15.8) sessions of routine CBT, we modeled both group-level and individual symptom networks of temporal change. First, we tested whether the group-level network structure could be generalized to individuals without significant loss of information using the ergodicity information index (EII). Second, we used information-theoretic mixture clustering (ITMC) to empirically detect subgroups for which structural ergodicity holds. The baseline characteristics of these subgroups were then compared using regularized multinomial regression.</p> Results <p>The EII indicated structural non-ergodicity, showing that population-level symptom change networks could not adequately represent individual networks. ITC identified five structurally distinct subgroups, two of which were sufficiently large to confirm structural ergodicity. Subgroups differed meaningfully in dynamic network structures and baseline clinical profiles, with one characterized by higher depression and somatic concerns and another by anxious-avoidant features.</p> Conclusions <p>The empirical testing routine for structural ergodicity used in this paper offers a promising approach to examining the possible loss of information when generalizing from group network structures to individual patients. This approach has practical implications for routine outcome monitoring, enabling more refined and individualized assessment strategies while maintaining the scalability of nomothetic frameworks.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Must We Always Go Idiographic?

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年11月11日

查看全文
  • Tim Kaiser,
  • Philipp Herzog,
  • Eva-Lotta Brakemeier,
  • Hudson Golino

摘要

Purpose

While psychological change processes are increasingly assumed to be “non-ergodic”, prompting a shift toward idiographic approaches, the assumption of ergodicity is often accepted a priori rather than tested empirically. This study investigates the validity of this assumption in the context of symptom dynamics during cognitive-behavioral therapy (CBT), focusing on the equivalence of network structures, i.e., structural ergodicity.

Methods

Using a sample of weekly transdiagnostic symptom assessments from N = 230 patients who received an average of 39.5 (SD = 15.8) sessions of routine CBT, we modeled both group-level and individual symptom networks of temporal change. First, we tested whether the group-level network structure could be generalized to individuals without significant loss of information using the ergodicity information index (EII). Second, we used information-theoretic mixture clustering (ITMC) to empirically detect subgroups for which structural ergodicity holds. The baseline characteristics of these subgroups were then compared using regularized multinomial regression.

Results

The EII indicated structural non-ergodicity, showing that population-level symptom change networks could not adequately represent individual networks. ITC identified five structurally distinct subgroups, two of which were sufficiently large to confirm structural ergodicity. Subgroups differed meaningfully in dynamic network structures and baseline clinical profiles, with one characterized by higher depression and somatic concerns and another by anxious-avoidant features.

Conclusions

The empirical testing routine for structural ergodicity used in this paper offers a promising approach to examining the possible loss of information when generalizing from group network structures to individual patients. This approach has practical implications for routine outcome monitoring, enabling more refined and individualized assessment strategies while maintaining the scalability of nomothetic frameworks.