Classical test theory | A theory about test scores, which postulates that a person’s test score can be modeled as the sum of two unobservable variables, a true score and an error score where with certain assumptions a true score can be shown to be the expected score across parallel test forms. A limitation of the classical test theory model is that both person scores and item statistics are dependent on the test and the sample of examinees, respectively. |

Item response theory | A modern statistical theory for test development and scoring that improves the understanding of both individual and item performance. It is based on the concept that both test items and individuals can be simultaneously described in terms of their standing on, or relationship to, one or more ability/knowledge level scales. An individual is characterized by his or her position on the scale, and an item can be characterized by the point on the scale indicating the trait or ability level at which a person would have a 50% chance of responding correctly to the item. |

Examination blueprint | A table of specifications based on the knowledge domain of family medicine, which is weighted by a judged frequency of use and clinical importance. It is used to determine the scope and depth of material covered in the examination process. |

Equating | A statistical process by which the score scale on one test is converted to the score scale of another test so that the scores are equivalent or parallel. |

Certifiably competent | A description of a candidate who has scored at or above the cut score level on a certification examination. |

Cut score | A specified point on a score scale at or above which a candidate passes and below which a candidate fails. |

Dimensions/orders | A categorization of the domain of knowledge in family medicine, which serves as an organizing agent for the content blueprint. |

Adaptive testing | A sequential form of testing in which the order of items presented is determined by the examinee’s response to previous items. |