Computerized adaptive and multistage testing with R

Abstract

Adaptive testing is, by now, a well-established procedure in educational and psy- chological assessment. Early adopters in the educational field in the USA were the licensing exam (NCLEX/CAT) of the National Council of State Boards of Nursing and the Graduate Record Examination (GRE). But at the time of this writing, the trend of using adaptive testing is worldwide. For instance, in the Netherlands, sev- eral adaptive tests are available for the compulsory nationwide final assessments in primary education. Further, also large-scale international educational surveys have implemented adaptive assessments; the Organisation for Economic Co-operation and Development (OECD) Program for International Student Assessment (PISA) and Program for the International Assessment for Adult Competencies (PIAAC) are prominent examples. Adaptive testing has also reached fields beyond educational assessment, such as epidemiology (the Patient-Reported Outcomes Measurement Information System (PROMIS) project for health assessment is an example), and organizational assessment (for supporting selection and promotion decisions). Most of these applications are based on adaptive selection of individual items. A more recent development is towards adaptive selection of testlets. This development is mainly motivated by the ease with which content control can be supported and generally goes under the name of multistage testing.