Nonparametric goodness-of-fit tests for the rasch model

Ponocny, Ivo

doi:10.1007/bf02294444

Cited by 81 publications

(104 citation statements)

References 20 publications

Supporting

Mentioning

101

Contrasting

Order By: Relevance

“…The result of the Fig. 2 Correlation matrix presented in the study by Koch (2006) non-parametric T11 test (p value = .824) (Ponocny 2001) shows that this procedure was successful.…”

Section: Intended Information For the High-stakes Assessment And Testmentioning

confidence: 87%

Prospective educators as consumers of empirical research: an authentic assessment approach to make their competencies visible

Wiethe-Körprich¹,

Bley²

2017

Empirical Res Voc Ed Train

View full text Add to dashboard Cite

Educators in vocational education and training (VET) should act as consumers of empirical researchUndertaking empirical research stimulates profitable innovation (Egeln et al. 2002). Natural scientists, for instance, have long taken it for granted that they should base their practical actions on current scientific research. In the healthcare sector, for example, hardly anybody wants to be treated by a doctor who refers to outdated research findings (Jahed et al. 2012). In many professions it is now common that practitioners are obliged to know about the latest relevant research results. Educators too have to be familiar with principles of empirical research in so far as they are able to reflect and to Abstract Background: Educators today have to be able to make current empirical research results usable for everyday practice. Consequently, there are increasing endeavors to develop and assess competencies in consuming empirical research (CCER) on an academic level. However, problems with regard to recruiting and motivating test participants-rooted in the prevalence of low-stakes testing conditions-could limit confidence in the validity of the findings. The current study presents a structure and proficiency level modeling for CCER under high-stakes conditions. Method:The sample comprises N = 155 bachelor students of Human Resource Education and Management. The assessment design of the 26 items complied with demanding standards for designing tests (such as Evidence-Centered Design and authenticity). Results:The results are as follows: (1) We were able to confirm our expected structural model which consists of two dimensions ('conceptual competencies' and 'statistical competencies') instead of one overarching dimension. (2) The test items are of a high quality. (3) Three levels of CCER could be defined according to two task characteristics (cognitive processes and complexity) which explain nearly 100% of the prospective educators' CCER abilities. Conclusion:The results of the study show that we succeeded in designing a reliable and valid test instrument for assessing (prospective) VET-educators' competencies in consuming empirical research. Keywords: High-stakes testing, Item-response-theory, Research competenciesOpen Access © The Author(s) 2017. This article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made. Wiethe-Körprich and Bley Empirical Res Voc Ed Train (2017) Page 2 of 26 Wiethe- Körprich and Bley Empirical Res Voc Ed Train (2017) 9:8 critically question the findings of scientific research. Correspondingly, it has been suggested that evidence-based practice should prompt educational professionals to be aware of recent advances in their area of work (Darling-Hammond and Bransford 2...

show abstract

“…The result of the Fig. 2 Correlation matrix presented in the study by Koch (2006) non-parametric T11 test (p value = .824) (Ponocny 2001) shows that this procedure was successful.…”

Section: Intended Information For the High-stakes Assessment And Testmentioning

confidence: 87%

Prospective educators as consumers of empirical research: an authentic assessment approach to make their competencies visible

Wiethe-Körprich¹,

Bley²

2017

Empirical Res Voc Ed Train

View full text Add to dashboard Cite

show abstract

“…MCMC type methods for sampling from the 14 psychometrika uniform distribution of zero-one tables with fixed marginal sums have been considered by Connor and Simberloff (1979), Besag and Clifford (1989), Roberts and Stone (1990), and Rao, Jana and Bandyopadhyay (1996). Ponocny (2001) uses a variation of the MCMC method. He generates zero-one matrices with fixed marginal sums according to a Markov chain but the limiting distribution of the chain is not the desired uniform distribution.…”

Section: Monte Carlo Methodologymentioning

confidence: 99%

“…A second practical motivation for Rasch's testing approach is that one can employ an intuitive test statistic with power against an alternative of interest whose asymptotic distribution may be unknown. Ponocny (2001) develops many such test statistics, including ones that test local stochastic independence, group differences, item discrimination, and dynamic test behavior. In Section 6, a test statistic for testing unequal item discrimination, whose power is considerably greater than other available test statistics is formulated.…”

Section: Assumptions Of the Rasch Model And Rasch's Approach To Testingmentioning

confidence: 99%

“…Most of the tests that have been developed rely on asymptotic theory for their validity (see the survey by Glas & Verhelst, 1995; for some exceptions, see Molenaar, 1983;Ponocny, 2001). Rasch was not very comfortable with any form of asymptotic theory (Andersen, 1995).…”

Section: Introductionmentioning

confidence: 98%

“…Rasch's testing approach is difficult to implement because no good combinatorial methods or analytical methods are known for deriving the distributions of test statistics under this complicated null conditional distribution (Snijders, 1991). Ponocny (2001) has recently developed a Monte Carlo method for approximating the distribution of test statistics in Rasch's approach along with an extensive set of test statistics for use 12 psychometrika in Rasch's approach. Ponocny demonstrates the power and flexibility of his procedure through an empirical example in which his procedure identifies the existence of subscales, violations of sufficiency of the raw score, and learning processes.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Exact tests for the rasch model via sequential importance sampling

Chen

Small

2005

Psychometrika

View full text Add to dashboard Cite

show abstract

R asch Modeling

Fischer¹

2005

Encyclopedia of Statistics in Behavioral Science

View full text Add to dashboard Cite

The Rasch Model is an item response model for ability or achievement tests with dichotomous item response format. Basic properties of the model are sufficiency of the raw score with respect to the person parameter and the so‐called ‘ specific objectivity ’ of comparisons between items and between persons. The model guarantees measurements with interval scale properties, except for very special cases where the scale becomes an ordered metric scale. The best‐founded approach to the estimation of the parameters of the model and to testing of fit or of other hypotheses about the parameters is the conditional likelihood method. Extensions of the model allow, for instance, to formalize structural hypotheses about the item parameters, to measure and test hypotheses about change, and to deal with scales composed of items with more than two ordered response categories.

show abstract

Nonparametric goodness-of-fit tests for the rasch model

Abstract: nonparametric tests, IRT, goodness-of-fit tests, Rasch model, exact tests,

Cited by 81 publications

References 20 publications

Prospective educators as consumers of empirical research: an authentic assessment approach to make their competencies visible

Prospective educators as consumers of empirical research: an authentic assessment approach to make their competencies visible

Exact tests for the rasch model via sequential importance sampling

R asch Modeling

Contact Info

Product

Resources

About