Figures

This page collects the central conceptual figures for the project.

Figure 1 – Core augmentation principle

flowchart LR
    A["Validated psychopathology questionnaire"] --> B["Item responses / symptom profile"]
    B --> C["Selection of personally relevant items"]
    C --> D["Assessment of perceived causal relations"]
    D --> E["Person-specific perceived causal network"]
    E --> F["Representativeness"]
    E --> G["Reliability"]
    E --> H["Incremental validity"]
    E --> I["Clinical / subjective utility"]

Figure 1. Conceptual overview of the proposed augmentation approach. After completion of a validated psychopathology questionnaire, personally relevant items are selected and used as nodes in a brief assessment of perceived causal relations. The resulting person-specific perceived causal network is then evaluated with respect to representativeness, reliability, incremental validity, and potential clinical utility.

Figure 2 – From questionnaire items to a person-specific network

flowchart LR
    subgraph Q["Questionnaire responses"]
        Q1["Low mood = 2"]
        Q2["Sleep problems = 3"]
        Q3["Worrying = 2"]
        Q4["Concentration problems = 2"]
        Q5["Social withdrawal = 1"]
        Q6["Irritability = 0"]
    end

    Q --> S

    subgraph S["Selected relevant items"]
        S1["Sleep problems"]
        S2["Worrying"]
        S3["Low mood"]
        S4["Concentration problems"]
        S5["Social withdrawal"]
    end

    S --> N["Perceived causal ratings"]
    N --> R["Person-specific network"]

Figure 2. Illustration of how conventional questionnaire responses can be transformed into a person-specific causal model. After questionnaire completion, a subset of personally relevant items is selected. Participants then rate perceived causal relations among these selected problems, resulting in an individualized network representation.

Figure 3 – Incremental-validity framework

flowchart TB
    subgraph B["Base model"]
        B1["Questionnaire total / subscale scores"]
        B2["Item profile"]
        B3["Baseline outcome"]
    end

    B --> O1["Outcome at follow-up"]

    subgraph A["Augmented model"]
        A1["Questionnaire total / subscale scores"]
        A2["Item profile"]
        A3["Baseline outcome"]
        A4["PCN density"]
        A5["Mean causal strength"]
        A6["Out-strength dispersion"]
    end

    A --> O2["Outcome at follow-up"]

    O1 --> C["Incremental value?"]
    O2 --> C

Figure 3. Incremental-validity framework for testing the added value of perceived causal network (PCN) augmentation. The central comparison contrasts a base model using conventional questionnaire information with an augmented model that additionally includes prespecified network characteristics.

Figure 4 – Same symptoms, different networks

flowchart LR
    subgraph P["Very similar symptom profile"]
        P1["Low mood = 2"]
        P2["Sleep problems = 2"]
        P3["Worrying = 2"]
        P4["Concentration problems = 2"]
    end

    P --> A1
    P --> B1

    subgraph A["Person A"]
        A1["Sleep problems"]
        A2["Low mood"]
        A3["Concentration problems"]
        A4["Worrying"]

        A1 --> A2
        A1 --> A3
        A4 --> A1
    end

    subgraph B["Person B"]
        B1["Worrying"]
        B2["Sleep problems"]
        B3["Low mood"]
        B4["Concentration problems"]

        B1 --> B2
        B1 --> B3
        B1 --> B4
    end

Figure 4. Illustration of person specificity. Even when participants show highly similar conventional symptom profiles, they may report markedly different perceived causal structures. In this example, sleep problems are central in Person A, whereas worrying acts as the dominant driver in Person B.

Figure 5 – Proposed survey workflow

flowchart TD
    A["Consent and eligibility"] --> B["Demographics"]
    B --> C["Questionnaire 1"]
    C --> D["Item selection"]
    D --> E["PCN ratings"]
    E --> F["Network visualization"]
    F --> G["Feedback / representativeness ratings"]

    G --> H["Questionnaire 2"]
    H --> I["Item selection"]
    I --> J["PCN ratings"]
    J --> K["Network visualization"]
    K --> L["Feedback ratings"]

    L --> M["Questionnaire 3"]
    M --> N["Item selection"]
    N --> O["PCN ratings"]
    O --> P["Network visualization"]
    P --> Q["Feedback ratings"]

    Q --> R["Optional global comparison"]
    R --> S["Optional retest / follow-up"]

Figure 5. Proposed survey workflow for the initial student validation study. Each questionnaire is followed by item selection, perceived causal network assessment, visualization, and immediate feedback ratings. A retest or follow-up can be added for reliability and predictive-validity analyses.