The Rasch Rating Model and the Disordered Threshold Controversy

Adams, Raymond J.; Wu, Margaret; Wilson, Mark

doi:10.1177/0013164411432166

Cited by 104 publications

(114 citation statements)

References 19 publications

Supporting

Mentioning

111

Contrasting

Unclassified

Order By: Relevance

“…The disordered threshold indicated issues with the discrimination of the item and could have caused problems with validity and reliability. Items with disordered parameter estimates need to be reviewed with regard to these issues (Adams, Wu, & Wilson, 2012;Andrich & Styles, 2004). This would improve the response categories conforming to the recommended Rasch Andrich thresholds even though the reliability of both person and item decreased.…”

Section: Discussionmentioning

confidence: 99%

Development and validation of a 6‐item Revised UCLA Loneliness Scale (RULS‐6) using Rasch analysis

Wongpakaran

Pinyopornpanish

Simcharoen

et al. 2020

British J Health Psychol

View full text Add to dashboard Cite

Objective. The UCLA Loneliness Scale, containing 20 items, is one of the commonly used loneliness scales. Some shorter versions have been developed using factor analysis. The study aimed to shorten the UCLA Loneliness Scale using Rasch and factor analysis methods and test the psychometric properties of the new scale.Methods. The full sample of the study included 719 respondents, divided into three subsamples (205, 324, and 190 for samples 1-3, respectively). The original, 20-item Revised UCLA Loneliness Scale (R-ULS) was shortened using 205 students (sample 1); the shortened scale was then validated for construct and concurrent validity with 324 students (sample 2) and 190 clinical participants (sample 3). Confirmatory factor analysis and Rasch analysis were used for construct validity. Convergent, discriminant, and concurrent validity were assessed by exploring the correlation with other psychological measurements.Results. In sample 1, the R-ULS was shortened to a 6-item scale (RULS-6) that fits the Rasch model. The RULS-6 met the criteria of unidimensionality and local independence without differential item functioning due to age and sex, and good targeting the clinical sample. Person Separation Index (PSI) reflected that reliability from the Rasch perspective was acceptable. However, collapsing categories 2 (sometime) and 3 (rarely) may be required in a clinical sample. When tested in samples 2 and 3, the RULS-6 fits the Rasch measurement model. Convergent and discriminant validity were demonstrated with interpersonal problems and attachment scales. As expected, a positive correlation was found between RULS-6 and anxiety, depression subscale, interpersonal difficulties, and somatization subscales denoting concurrent validity. Cronbach's alpha of the RULS-6 was good (.83).Conclusion. Using Rasch analysis, the proposed RULS-6 constituted a 70% reduction of the number of original items, yet preserved the psychometric properties in independent samples of students and psychiatric outpatients.

show abstract

Section: Discussionmentioning

confidence: 99%

Development and validation of a 6‐item Revised UCLA Loneliness Scale (RULS‐6) using Rasch analysis

Wongpakaran

Pinyopornpanish

Simcharoen

et al. 2020

British J Health Psychol

View full text Add to dashboard Cite

show abstract

“…The dimensionality study provided by Amburgey and Thoman (2012) found measurement issues in item 9 as well, where the item did not load as a signifi cant predictor for ecological concern in a confi rmatory factor analysis. Another hypothesis can be found in Adams, Wu and Wilson (2012), where it is discussed that the lack of frequencies in response categories within the rating scale might bias fi tting measures -and responses on lower categories were scarce for both items 3 and 9. Further investigation is needed with a balanced sample through response categories.…”

Section: Discussionmentioning

confidence: 99%

Propriedades psicométricas da versão brasileira da escala de novo paradigma ecológico: revisada

et al. 2016

View full text Add to dashboard Cite

The New Ecological Paradigm (NEP-R) is one of the hallmarks for the measurement of ecological attitudes. Attention has been given to its psychometric properties during the last ten years, but no analyses have been conducted until now using an Item Response Theory (IRT) approach. With this purpose, a sample of 410 recruited responders answered the NEP-R's 15 items in Brazilian Portuguese. An item analysis was conducted considering: dimensionality; reliability; discrimination; item information; and item/person hierarchy. The results for the principal component analysis of standardized residues supports a unidimensional structure. Results for item fi t coeffi cients (INFIT and OUTFIT) indicate that the items have good predictive power for the response patterns in the latent trait continuum representing ecocentrism. In turn, the results for person discrimination and item hierarchy point out that the scale is restricted as to the levels of diffi culty it represents. The fi ndings include that the most revealing items for environmental attitudes in a Brazilian sample consist of the set of items representing the "limits of growth" facet.

show abstract

“…The ordering of the PCM's categories is independent of the ordering of the thresholds and has to be assumed prior to data analysis (Masters, 1988). For a more detailed formal treatment of the distinction between the ordering of the response categories and the ordering of the thresholds, see Adams et al (2012).…”

Section: Relationship Between Category Probabilities and Threshold Ormentioning

confidence: 99%

Reversed Thresholds in Partial Credit Models

Wetzel

Carstensen

2014

Assessment

View full text Add to dashboard Cite

ArticleOrdered rating scales are widely used in the assessment of personality, attitudes, and other latent variables. For example, in the Revised NEO Personality Inventory (NEO-PI-R; Costa & McCrae, 1992), participants respond on a 5-point Likert-type scale with the options strongly disagree, disagree, neutral, agree, and strongly agree. Another example for an ordered rating scale is the response categories never, sometimes, often, and always. With ordered rating scales the underlying assumption is that endorsing a higher response category implies a higher trait level.Models in the framework of item response theory (IRT) such as the Rasch model (Rasch, 1960) for dichotomous items or the partial credit model (PCM;Masters, 1982) for polytomous items define the probability of a response in a certain category as a function of the respondent's latent trait level and item characteristics. In the Rasch model, only one item parameter is estimated, namely, the difficulty of the item. In modeling responses from ordered rating scales according to the PCM, the "difficulty" of each of the response categories needs to be taken into account. This is done using threshold parameters that are defined as the point on the latent trait continuum where the response probability for two adjacent response categories is equal. Thus, for a 5-point scale, we have four threshold parameters. To illustrate, Figure 1A shows category probability curves for an item with a 5-point scale. These curves represent the probability of endorsing each of the five categories conditional on the latent trait level depicted on the x-axis. The four thresholds are included as perpendicular lines. Threshold 1, which is the threshold between the categories strongly disagree and disagree, is located at about −2.9 logits. Thus, respondents with trait levels of exactly −2.9 have equal probabilities of endorsing strongly disagree and disagree. Respondents with trait levels below −2.9 have the highest probability of endorsing strongly disagree, whereas respondents with trait levels above −2.9 to −0.8 (where threshold 2 is located) have the highest probability of endorsing disagree. This definition and interpretation of threshold parameters also holds for extensions of the PCM to two-parameter logistic models such as the generalized PCM (GPCM; Muraki, 1992). In the Rasch model and the PCM, it is assumed that all items have the same item discrimination. Two-parameter logistic models for dichotomous or polytomous items (e.g., GPCM) relax this assumption and estimate a discrimination parameter for each item in addition to the difficulty (threshold) parameters. While most of the following discussion will focus on the one-parameter PCM, some of the analyses will also be reported for the GPCM to show that the results generalize to models with more than one parameter. AbstractWhen questionnaire data with an ordered polytomous response format are analyzed in the framework of item response theory using the partial credit model or the generalized partial credit model, revers...

show abstract

The Rasch Rating Model and the Disordered Threshold Controversy

Cited by 104 publications

References 19 publications

Development and validation of a 6‐item Revised UCLA Loneliness Scale (RULS‐6) using Rasch analysis

Development and validation of a 6‐item Revised UCLA Loneliness Scale (RULS‐6) using Rasch analysis

Propriedades psicométricas da versão brasileira da escala de novo paradigma ecológico: revisada

Reversed Thresholds in Partial Credit Models

Contact Info

Product

Resources

About