Rasch Consultants

The primary focus of Measurement Research Associates, Inc. is to provide personalized examination development, analysis and reporting services. Services include the following. Traditional statistical analyses are performed for written multiple choice examinations. Criterion referenced standard setting methods (e.g., bookmark) can be explained and implemented.

Latent trait analysis, using the Rasch model, offers additional information about the performance of written multiple choice examinations. Using the Rasch model, examination equating ensures that the same standard is used across examination administrations. Multi-facet Rasch analysis has been used successfully for oral, practical, case and other examinations that requir e the intervention of a rater. The impact of all facets of the examination are accounted for, including rater severity, item or case difficulty, and skill difficulty. The reliability of the examinati on is enhanced, because the error of measurement is decreased. Thus, multi-facet examinations can be shown to be reliable, as well as, valid. Computerized adaptive testing is a method of using the power of the computer to construct individualized tests for each candidate, thus decreasing the error of measurement more quickly, so that tests can be shorter without sacrificing precision or confidence in the pass/fail decisions. Test security is potentially improved by using adaptive testing.

If you are interested in learning more, please contact Mary E. Lunz, Ph.D. at 312-822-9648, or e-mail to MeasResInc@aol.com

Graduate of MESA, University of Chicago, 1996
Psychometrician with Educational Data Systems, Morgan Hill, CA
Owner, Pythias Consulting, LLC, San Jose, CA
email: pythiasconsulting ~at~ gmail.com

I apply psychometric models to audio listening tests (Moulton Laboratories and Lucent Technologies), program evaluation (Reading First), large-scale assessment (CHSPE), data mining (profiling travelers), and district assessments (equating). My specialty is Rasch models -- including Many-Facet Rasch Models -- and multidimensional models developed under the name "NOUS" in accordance with Rasch requirements for "objectivity." My company, Pythias Consulting, markets NOUS software for specialty applications.

One such application is called EdScale, developed for Educational Data Systems. It converts test-data from district assessments administered through the school-year into California Standards Test (CST) scale scores. EdScale measures student growth within and across years and helps district administrators predict whether schools are on track to meet NCLB requirements.

I have been one of the primary architects of the California Reading First evaluation for the past 7 years, using a 3-Facet Rasch Model to measure implementation at each of over 850 schools and relating these measures to achievement gain. We have been able to demonstrate a strong Reading First effect in California (in contrast to a widely cited national study).

My work can be accessed at www.eddata.com

Consultant in Rasch, logistic models and classical models for educational evaluation and biostatistics. I give technical support from the design of a test to item analysis. I produce the family of KALT programs (including item banking, classical item analysis, on-line testing including Rasch model and criterion-referenced test and item analysis) and offer technical support and training (test design, Rasch analysis, item design and analysis). An interesting service that I am offering since 2000 is the "Analysis of quality of a test" according to international standards (among the areas that we evaluate are: validity, objectivity, reliability, logistics and application, fair test practices, research, etc.) , in this case the institution provides information and evidences of a set of elements of the test and we provide a technical report indicating the areas that are correct and the areas needing an improvement according to the standards. I can offer my services in English, Spanish and French.
Web page (Spanish and English): www.ieesa-kalt.com