Better Rating Scale Scores with Information–Based Psychometrics

Ramsay, J. O.; Li, Juan; Wiberg, Marie

doi:10.3390/psych2040026

Cited by 10 publications

(7 citation statements)

References 19 publications

(26 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We wanted to argue that the choice latent trait metric is arbitrary in IRT models, and the θ or the δ metric can be both useful in applications. We tend to prefer bounded trait metrics in applications because it seems more challenging to interpret the possibility of unbounded negative and positive trait values of θ [33]. However, we would prefer the true score metric τ or the rank score ρ over δ.…”

Section: Discussionmentioning

confidence: 99%

About the Equivalence of the Latent D-Scoring Model and the Two-Parameter Logistic Item Response Model

Robitzsch¹

2021

Preprint

View full text Add to dashboard Cite

This article shows that the recently proposed latent D-scoring model of Dimitrov is statistically equivalent to the two-parameter logistic item response model. An analytical derivation and a numerical illustration are employed for demonstrating this finding. Hence, estimation techniques for the two-parameter logistic model can be used for estimating the latent D-scoring model. In an empirical example using PISA data, differences of country ranks are investigated when using different metrics for the latent trait. In the example, the choice of the latent trait metric matters for the ranking of countries. Finally, it is argued that an item response model with bounded latent trait values like the latent D-scoring model might have advantages for reporting results in terms of interpretation.

show abstract

Section: Discussionmentioning

confidence: 99%

About the Equivalence of the Latent D-Scoring Model and the Two-Parameter Logistic Item Response Model

Robitzsch¹

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Data analysis observed that the relationship between physical activity and fall occurrence did not follow a linear relation. The function (quartic) which best t this relation (showing high R 2 , signi cant p-value, and high F-statistic) was determined using the curve estimation regression, and the probability of falling was calculated depending on the amount of physical activity i.e., metabolic expenditure [46,47]. The sensitivity and speci city of this model were determined using ROC analysis [48].…”

Section: Discussionmentioning

confidence: 99%

Dynamic Performance-exposure Algorithm for Falling Risk Assessment and Fall Prevention in Community-dwelling Older Adults

Pereira

Rosado

Almeida

et al. 2021

Preprint

View full text Add to dashboard Cite

Background: Several models and algorithms were designed to identify older adults at risk of falling supported on an intrinsically and extrinsically traditional approach. However, the dynamic interaction between multiple risk factors for falls must be considered. The present study aimed to design a dynamic performance-exposure algorithm for falling risk assessment and fall prevention in community-dwelling older adults.Methods: The study involved 1) a cross-sectional survey assessing retrospective falls, performance-related risk factors for falls (sociodemographic such as gender and age, cognitive, health conditions, body composition, physical fitness, and dual-task outcomes), exposure risk factors (environmental hazards and (in)physical activity), and performance-exposure risk factors (affordance perception), and 2) follow-up survey assessing prospective falls. Participants were Portuguese community dwellings (≥ 65 years). Data were reported based upon descriptive statistics, curve estimation regression, binary logistic regression, and ROC curve.Results: The selected and ordered outcomes included in the algorithm and respective cutoffs were: (1) falls in previous year (high risk: n>1, moderated: n=1, low risk: n=0); (2) health conditions (high risk: n >3, moderated: n=3, low risk: n<3); (3) multidimensional balance (high risk: score <32 points, moderated risk: 32 points ≤ score ≤33 points, low risk: score>33); (4) lower body strength (high risk: rep/30s< 11, moderated risk: 11≤ rep/30s ≤14, low risk: rep/30s >4); (5) perceiving action boundaries (high risk: overestimation bias, moderated risk: not applied, low risk: underestimation bias); (6) fat body mass (high risk: % fat >38, moderated risk: 37≤ % fat ≤38, low risk: % fat <7); (7) environmental hazards (high risk: n>5, moderated risk: n=5, low risk: n<5); (8) rest period (high risk: hours/day >4.5, moderated risk: 4≤ hours/day ≤4.5, low risk: hours/day <4); (9) physical activity metabolic expenditure (high risk: MET-min/week <2300 or >5200, moderated risk: 2300≤ MET-min/week <2800, low risk: 2800≤ MET-min/week ≤5200).Conclusions: Results demonstrated a dynamic relationship between older adults’ performance capacity and the exposure to falls opportunity, supporting the build algorithm’s conceptual framework. Fall prevention measures should consider the above factors that most contribute to the individual risk of falling, relative weights, and their distance from low-risk value, as proposed in the dynamic algorithm.

show abstract

“…The focus of the evaluation in the IRT analysis is on the individual items, enabling the evaluation of different item parameters (DeVellis, 2017 ; Tavakol et al, 2014 ) and providing an informative way to analyse composite scales consisting of several categorical items that are summarized into a total score (Wellhagen et al, 2021 ). The TestGardener software applies modern statistical methods to produce accurate estimates of respondent characteristics using full data and enables item‐level analysis and full distractor (incorrect response) analysis (Li et al, 2019 ; Ramsay et al, 2020 ). TestGardener can be used to evaluate problems with items and help test developers decide whether to rewrite items to clarify ambiguous wording or to modify incorrect options to make them more plausible.…”

Section: Methodsmentioning

confidence: 99%

“…TestGardener can be used to evaluate problems with items and help test developers decide whether to rewrite items to clarify ambiguous wording or to modify incorrect options to make them more plausible. The software provides a visual S‐shaped logistic curve (item characteristic curve [ICC]) in which different items' response options can be graphically analysed (Li et al, 2019 ; Ramsay et al, 2020 ). It is essential that the items are unambiguous (i.e.…”

Section: Methodsmentioning

confidence: 99%

See 1 more Smart Citation

Development and psychometric evaluation of the Emotional Intelligence Test (EMI‐T) for social care and healthcare student selection

Pienimaa

Talman

Vierula

et al. 2022

Journal of Advanced Nursing

View full text Add to dashboard Cite

Aim To develop an emotional intelligence (EI) test and evaluate its psychometrics for social and healthcare student selection. Design A cross‐sectional methodological design. Methods The test was developed based on a systematic review and focus group interviews. Content validity was evaluated with expert panels, and preliminary psychometrics with two pilot studies. Descriptive statistics, correlations and item response theory were used. Data Sources Search was conducted in six databases 2018. Focus group interviews were conducted with educators and professionals in 2019. Expert panels with doctoral students, researchers and educators were conducted in 2020. Pilot tests with students were conducted 2020–2021. The developed test was administered to 4808 applicants 2021. Results The test included four subscales. Correlations support the test's theoretical structure. The items were mainly easy. Conclusion The test assesses EI objectively and comprehensively. The item‐level distractor analysis can be used for further test development. Impacts Social care and healthcare students engage in clinical practice early in their studies, and these environments can be emotionally challenging. Assessing EI in student selection with adequate test can help the institutions of higher education to select the students with required abilities to succeed in the studies. The assessment of EI during student selection also provides information higher education institutions could use to develop and provide support interventions. The results may also encourage practice placements to include EI elements as learning objective. The results of this study and especially the use of IRT and detailed distractor analysis to evaluate the psychometric properties of EMI‐T can benefit researchers and educators that develop or evaluate objective assessment tools with multiple choice questions. Implications for the profession and/or patient care Emotional intelligence is important for students to enable professional interaction.

show abstract

Better Rating Scale Scores with Information–Based Psychometrics

Cited by 10 publications

References 19 publications

About the Equivalence of the Latent D-Scoring Model and the Two-Parameter Logistic Item Response Model

About the Equivalence of the Latent D-Scoring Model and the Two-Parameter Logistic Item Response Model

Dynamic Performance-exposure Algorithm for Falling Risk Assessment and Fall Prevention in Community-dwelling Older Adults

Development and psychometric evaluation of the Emotional Intelligence Test (EMI‐T) for social care and healthcare student selection

Contact Info

Product

Resources

About