1. Because it is not feasible to create and administer tests containing every possible item related to a domain, trait, or skill (it would take a lifetime to create this test. let alone take it!) we must create tests consisting of a representative :] o of items from the domain. Hide Feedback 2. Reliability analyses estimate how much l[:]lo there is in a score from a shorter test that is meant to mirror the longer (impossible-tocreate) test containing all possible items. Hide Feedback