Quantifying and addressing the impact of measurement error in network models

Ron, Jill de; Robinaugh, Donald J.; Fried, Eiko I.; Pedrelli, Paola; Jain, Felipe A.; Mischoulon, David; Epskamp, Sacha

doi:10.1016/j.brat.2022.104163

Cited by 24 publications

(13 citation statements)

References 49 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The second step of the analysis was to evaluate item redundancy. A psychological network should be composed of unique causal components, that is, each item should measure a unique behavior, cognition or emotion (de Ron et al, 2022). Redundancy was evaluated with the weighted topological overlap (wTO) statistic (Zhang & Horvath, 2005) with adaptive alpha (Pérez & Pericchi, 2014).…”

Section: Discussionmentioning

confidence: 99%

Development and validation of the COVID-19 Impact Scale in Australia

McCormick¹,

Sethi²,

Haag³

et al. 2023

Preprint

View full text Add to dashboard Cite

The COVID-19 pandemic significantly impacted and continues to impact the health and well-being of Australian adults. However, there has been no instrument validated to comprehensively measure how the COVID-19 pandemic impacted adults in Australia across several domains (e.g. fear of COVID-19, attitudes towards vaccination, impact of lockdowns on psychosocial ill-being). The current study conducted a rigorous psychometric process to develop and validate an instrument to measure the impact of the COVID-19 pandemic in Australia, the COVID-19 Impact Scale (CIS). An initial pool of 30 items was developed based on a review of the literature and input from a panel of experts including psychologists, epidemiologists, and public health experts, among others. The study used network psychometrics to examine the psychometric properties of: (1) item score distributions; (2) item redundancy; (3) dimensionality; (4) model fit; (5) measurement invariance; (6) reliability; and (7) criterion validity. The 18-item CIS displayed a three-dimensional structure (“Fear”, “Attitudes”, “Psychosocial ill-being”), measurement invariance, good reliability, and criterion validity. The instrument is available to be used by Australian researchers and implemented to evaluate public policies, adapted for future pandemics, or used internationally.

show abstract

Section: Discussionmentioning

confidence: 99%

Development and validation of the COVID-19 Impact Scale in Australia

McCormick¹,

Sethi²,

Haag³

et al. 2023

Preprint

View full text Add to dashboard Cite

show abstract

“…If the items were similar in content (due to measuring the same “construct”) but conceptually different enough to preclude item exclusion, the items were summed together generating a composite item (i.e. “super item”) [63] .…”

Section: Methodsmentioning

confidence: 99%

Development and validation of an HPV infection knowledge assessment scale among Aboriginal and Torres Strait Islander Peoples

et al. 2023

View full text Add to dashboard Cite

“…The regularised partial correlation networks were estimated using EBICglasso regularisation with a default Extended Bayesian Information Criterion tuning parameter of 0.5 18 . To avoid potential impacts of measurement error by performing network analyses with only individual items as nodes, 19 an initial factor‐level network structure (Network A) was computed using the original five CDRISC factors 7 plus the nine PHQ‐9 items. Following this, factors which were weakly associated with MDD in Network A were eliminated and the focus shifted to CDRISC items which tapped the remaining factors.…”

Section: Methodsmentioning

confidence: 99%

Network connectivity between psychological resilience and depression in prostate cancer patients

Sharpley,

Arnold,

Christie

et al. 2023

Psycho-Oncology

View full text Add to dashboard Cite

ObjectivesProstate cancer (PCa) patients often experience depression. One possible buffer against stress‐related depression is psychological resilience (PR), which has been described as heterogeneous in structure, like major depressive disorder (MDD). Although both of these constructs are central to understanding and assisting distressed PCa patients, no data have been reported on how they connect via network arrays at a component and symptom level. Such information has the potential to inform clinical practice with depressed PCa patients.MethodsUsing a cross‐sectional design, 555 PCa patients completed the Patient Health Questionnaire‐9 (PHQ‐9) and the Connor–Davison Resilience Scale (CDRISC). Data were analysed via network analysis.ResultsNetwork analysis indicated that various CDRISC factors interacted with different PHQ‐9 symptoms. For example, trust in one's instincts, tolerance of negative affect, and strengthening effects of stress (CDRISC) was associated with concentration problems and suicidal ideation (PHQ‐9); positive acceptance of change, and secure relationships (CDRISC) was linked to low self‐worth, anhedonia, fatigue/lethargy, motor problems, depressed mood, and concentration and appetite problems (PHQ‐9). Similarly heterogeneous associations were found between individual CDRISC items and PHQ‐9 symptoms. Network analytic figures depict both these sets of associations.ConclusionsAs well as confirming the heterogeneous nature of PR and MDD in PCa patients, these findings argue for the further development of ‘individualised’ medicine approaches when working with PCa patients and their experiences of depression.

show abstract

Quantifying and addressing the impact of measurement error in network models

Cited by 24 publications

References 49 publications

Development and validation of the COVID-19 Impact Scale in Australia

Development and validation of the COVID-19 Impact Scale in Australia

Development and validation of an HPV infection knowledge assessment scale among Aboriginal and Torres Strait Islander Peoples

Network connectivity between psychological resilience and depression in prostate cancer patients

Contact Info

Product

Resources

About