Speakers
Description
Visualization is an important task in communicating the results of an IRT model. From the many available diagrams, the PImap (aka Wright-Map) is useful for many purposes: Utilizing the fact that in IRT item and person parameters share a common scale, it juxtaposes both parameter domains (top part: person parameter distribution; bottom part: item parameter localizations). Thus, it not only allows for identifying threshold disorders at a glimpse, it also shows how the two domains agree with respect to a given sample. The traditional form of the PImap (e.\,g., as provided by eRm or the WrightMap packages) uses points for item difficulty(RM)/threshold(PCM) parameters. The present contribution introduces an extended version of the PImap, drawing not only the parameter locations but adds many further useful information of each item. An R package will be provided, allowing for drawing the extended diagrams for all commonly used IRT packages of R.
