About IRT tags

In Surpass, you can attach metadata (known as “tags”) to items and then use these tags to organise items when authoring content, creating tests, and viewing results. Tags are made up of “tag groups” and “tag values”.

Item Response Theory (IRT) tags are used to describe the likelihood of candidates answering items correctly in tests. In Surpass, IRT tags can be created as custom tag groups. You can set up custom tag groups in the Subjects screen of Setup. For more information, read Setting up tag groups.

IRT tags are used in LOFT tests to control the psychometric characteristics of a test form. You can create four custom IRT tag groups for LOFT tests: IRT_A, IRT_B, IRT_C and IRT_Domain. Refer to the following grid for more information on each tag group.

NOTE: You must create tag values for IRT_A, IRT_B, and IRT_C tag groups if you are creating LOFT tests using psychometric rules.
Tag Group Description
IRT_A IRT_A tag groups must contain numeric tag values greater than 0. IRT_A tag groups contain IRT_a tag values, which measure how an item distinguishes between candidates of varying ability.
IRT_B IRT_B tag groups must contain either positive or negative numeric tag values. IRT_B tag groups contain IRT_b tag values, which measure the difficulty of an item. 
IRT_C IRT_C tag groups must contain numeric tag values greater than or equal to 0, but less than one. IRT_C tag groups contain IRT_c tag values, which determine the likelihood a candidate of low ability will guess the correct answer in a test.

This tag group can contain tag values for which you want to generate an IRT theta value for statistical analysis.

EXAMPLE: This tag group could contain skills you want candidates to demonstrate in a test.

For more information about LOFT/Adaptive tests, read About LOFT/Adaptive tests.