RSP, Rasch scaling program. RSP analyzes dichotomous item responses with the Rasch model. These analyses are useful for test design, item banking, adaptive testing, item bias, change measurement, and educational achievement assessment. RSP provides several estimation procedures as well as global, item, and person fit indices. RSP can analyze up to 32 test forms, linked through common items or common person ability distributions. The maximum number of items per test is 96. The maximum number of different items is 600. There is no restriction on the number of persons taking a test, but no missing data is allowed within a test form. RSP permits conditional maximum likelihood estimates CMLE of item parameters that do not depend on any assumption about the distribution of person parameters. RSP also offers marginal maximum likelihood estimation under the assumption of one or more normal ability distributions. The RSP development team includes Cees Glas (CITO) and Jules Ellis (University of Nijmegen).
Keywords for this software
References in zbMATH (referenced in 2 articles )
Showing results 1 to 2 of 2.
- Feddag, M.L.; Mesbah, M.: Approximate estimation in generalized linear mixed models with applications to the Rasch model (2006)
- Béguin, A.A.; Glas, C.A.W.: MCMC estimation and some model-fit analysis of multidimensional IRT models (2001)