The sirens’ call in psychometrics: The invariance of IRT models

Asún, Rodrigo; Rdz-Navarro, Karina; Alvarado, Jesús M.

doi:10.1177/0959354317706272

Cited by 5 publications

(3 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This review provided the basis for our validation study. On the basis of the above results, the discussion for the use of classical test theory and item response theory in behavioral and social science [71], and the aim of our study (to adapt an already developed short scale), we decided to follow the classical test theory validation and the use of EFA. As there were many available validations providing different dimensions, we decided to explore the dimensions in this target group and confirm these factors in a larger study sample of carers.…”

Section: Resultsmentioning

confidence: 99%

Adapting the eHealth Literacy Scale for Carers of People With Chronic Diseases (eHeals-Carer) in a Sample of Greek and Cypriot Carers of People With Dementia: Reliability and Validation Study

Efthymiou¹,

Middleton²,

Charalambous³

et al. 2019

J Med Internet Res

View full text Add to dashboard Cite

BackgroundAs the population ages, many more people will be in need of long-term care. According to a recent report by Alzheimer's Disease International and the Karolinska Institute, 84% of people with dementia are cared for at home and 16% in nursing homes. Several Web-based interventions have been developed to assist the work of carers at home. Measuring the levels of electronic health (eHealth) literacy is of top priority to facilitate inclusion of this population and develop training programs to enhance eHealth literacy skills.ObjectiveThis study aimed to adapt the eHealth Literacy Scale (eHeals) for carers of people with dementia, who speak Greek as their native language and live in Greece and Cyprus, and to test the reliability and validity of the scale for carers.MethodsThe content validity of the eHealth Literacy Scale for Carers of People With Chronic Diseases (eHeals-Carer) was assessed with an expert panel (N=10). A descriptive study with face-to-face interviews among 101 primary carers of people with dementia was conducted. In addition to the eHeals-Carer to assess their perceived eHealth literacy, participants responded to a brief questionnaire regarding characteristics of internet use and provided sociodemographic data. The internal consistency of the tool and the construct validity via an exploratory factor analysis (EFA) were explored.ResultsThe Mean Item-Level Content Validity Index (CVI) and Scale-Level CVI Average was 0.93. The participants were mostly women (75.2%, 76/101), aged less than 60 years (67.3%, 68/101) with secondary education. The internal consistency was estimated at a Cronbach alpha of .83. Two factors were extracted from the EFA: information seeking questions 1 to 5 (factor 1) and evaluation questions 6 to 8 (factor 2).ConclusionseHeals-Carer is the first perceived eHealth literacy tool adapted for carers of people with dementia. The use of Web-based services available for carers could help them and improve the health care system in the long term. In Greece and Cyprus, there is a lack of services, and improving the digital skills of carers could provide them with the means to support themselves at home and improve care provision.International Registered Report Identifier (IRRID)RR2-10.2196/resprot.8080

show abstract

Section: Resultsmentioning

confidence: 99%

Adapting the eHealth Literacy Scale for Carers of People With Chronic Diseases (eHeals-Carer) in a Sample of Greek and Cypriot Carers of People With Dementia: Reliability and Validation Study

Efthymiou¹,

Middleton²,

Charalambous³

et al. 2019

J Med Internet Res

View full text Add to dashboard Cite

show abstract

“…We should stress that there is really nothing new about the idea of causal sensitivity analysis: fake-data simulation is often used in Bayesian data analysis to ensure that the model performs as expected, and, more importantly in this context, it is recognized as sometimes the only way to test if a model can reliably recover latent parameter values (Gelman et al, 2020). This is also how the impact of individual differences in the way of responding on test performance is sometimes studied (see, e.g., Asûn et al, 2017).…”

Section: What Both Analyses Illustratementioning

confidence: 99%

The general causal cumulative model of ordinal response

Paulewicz¹,

Blaut²

2022

Preprint

View full text Add to dashboard Cite

In this paper, we introduce the general causal cumulative model of ordinal response. The statistical part of this model is a new family of hierarchical or non-hierarchical generalized linear models that represent the distribution of the outcome as a thresholded latent distribution. Its defining feature is the new link functions, which are order-preserving in the sense that they allow for arbitrary effects in individual thresholds while preserving their order. We show how the model can be interpreted as a generalization of some Signal Detection Theory (SDT) models and some Item Response Theory models. We propose an approach to measurement of latent variables which seems to follow from the requirement that measurements should be interpreted in the context of a causal theory of the unobservable response process. In particular, we formulate a causal definition of measurement invariance that seems to match the examples of measurement bias found in the literature, and we show that the commonly accepted statistical definition is flawed; this leads to the recognition of the fundamental problem associated with item parameters and of the critical role of substantive theory and speculative reasoning. We illustrate how, by making use of the statistical properties of our model, this approach to measurement can be followed in practice, and we explain what kind of issues of a more technical nature it may help address and when. In particular, we identify the central causal assumption of SDT models, we show how this assumption can be tested, how introducing item parameters can lead to severely biased point and interval estimates of the target causal quantities, and how causal estimate bias can be accounted for by means of a causal sensitivity analysis.

show abstract

“…Although IRT approaches require larger samples and make stricter statistical assumptions than CTT, the item parameters they yield are -barring any differential item functioningpotentially subpopulation invariant (Embretson and Reise, 2000; for qualifications see Asun et al, 2017). This is a major advantage for the current research as this serves to produce test items and measurement scales that can function consistently across a range of samples, such as academic and work samples.…”

Section: Studymentioning

confidence: 99%

Burnout Stigma Inventory: Initial Development and Validation in Industry and Academia

May

Terman

Foster³

et al. 2020

Front. Psychol.

View full text Add to dashboard Cite

Although burnout is a risk factor for various negative mental and physical outcomes, its prevention is hampered by the stigma associated with burnout. The current research therefore reports on the initial development and validation of a novel measure of perceived burnout stigma. Study 1 (n = 318) describes the construction and initial evaluation of scale items derived from established mental health stigma and burnout scales. Study 2 (n = 705) then replicated the burnout stigma factor structure established in the initial study. Additionally, it evaluates relationships between occupational and school burnout stigma and indicators of mental health. Hierarchical multiple regressions showed that burnout stigma accounted for variance in depression, anxiety, and stress over and beyond that of burnout. Study 3 (n = 682) extended these findings via crosslagged and bidirectional models, demonstrating that burnout stigma predicted mental health indicators 6 weeks later. Study 4 (n = 717) supplemented earlier exploratory and confirmatory factor analyses using item response theory to further demonstrate that perceived burnout stigma is a unidimensional construct potentially applicable to both work and school settings. Overall, the current research resulted in an eight-item burnout stigma instrument (BSI-8) with excellent psychometric properties that predicts indicators of mental health.

show abstract

The sirens’ call in psychometrics: The invariance of IRT models

Cited by 5 publications

References 33 publications

Adapting the eHealth Literacy Scale for Carers of People With Chronic Diseases (eHeals-Carer) in a Sample of Greek and Cypriot Carers of People With Dementia: Reliability and Validation Study

Adapting the eHealth Literacy Scale for Carers of People With Chronic Diseases (eHeals-Carer) in a Sample of Greek and Cypriot Carers of People With Dementia: Reliability and Validation Study

The general causal cumulative model of ordinal response

Burnout Stigma Inventory: Initial Development and Validation in Industry and Academia

Contact Info

Product

Resources

About