10:832:346 Lecture Notes - Lecture 11: Program Evaluation, Criterion Validity, Face Validity
Document Summary
Create new/own data instruments and measures: 2. ) Equivalence reliability- extent to which different parallel versions or forms of a test measure the same concepts at the same level of difficulty, and produce similar results. Validity: example: patient satisfaction, generally refers to accuracy of a measure, 4 different types, 1. ) Creating a new measure: may be needed if no good existing measure exists that can be adapted to your needs, many steps needed so make sure you have, time, money, expertise needed to complete the task, checklist, 1. ) Define the subject matter or topics that will be covered and outline the content: consult literature and experts as needed, 3. ) Create item questions and item choices/answer choices: choices also requires knowledge about content/topic, 4. ) Select response choices for each question or item: easier for demographic information than data related to knowledge, attitudes, behavior, etc, 5. )