Thesis Defense - Practical Issues Related to Content Validity When Using Automated Test Assembly for Optimal Item Selection for a
Date: August 23, 2000
Time: 11:00 AM
Location: Commonwealth Hall, Room: 410
Aggie Butler, Ph.D. Candidate in Biomedical Science
Supervising Professor, Dr. Nira Herrmann
Practical Issues Related to Content Validity When Using Automated Test Assembly for Optimal Item Selection for a Medical Specialty Examination
Current developments in educational testing foster a renewed emphasis on evaluating the design of tests and the quality of test content. Empirical methods will be designed to gather data about the items used for automated test assembly and their related tasks that define the content domain for a medical specialty examination and to evaluate the ways in which the test matches its content and cognitive specifications. A fundamental purpose of the analysis is to identify attributes for items that are selected by content experts for an examination and to assess any latent characteristics associated with the item selection or the content areas in an examination. The attributes and characteristics provide descriptive information for each item, for each content area, and for the assessment as a whole.
These analyses provide applications for automated test assembly (ATA) that can be used in classifying items and item attributes for optimal item selection for parallel forms.
Every aspect of content for the medical specialty examination may not be amenable to multiple-choice testing, and the question of whether or not multiple-choice assessment tools might be suitable is an empirical question that also will be reviewed. Multiple-choice item development that focuses directly on analytical, logical, or critical thinking skills will be addressed.
Analysis of the relationship of test scores to a variable external to the examination will provide a source of validity evidence.