Abstract
This study was designed to examine the self-reported adult-caused testing irregularities experienced by schools taking the 2000 SAT-9 test in California. The population consisted of seven out of the eleven schools with such irregularities as identified by the California Department of Education. Each of these schools voluntarily reported the testing irregularities to the state department and confirmed the findings. Participation was voluntary. Due to the controversial nature and legal issues surrounding testing irregularities of the 2000 administration, some school districts either chose not to participate or could not participate due to a gag order imposed through litigation proceedings. Descriptive research with a qualitative case study design was utilized. Individual interviews were conducted with teachers, site and district testing coordinators, site principals, and county office assessment directors. Data were also retrieved through a review of archival records. The research results indicated that each case was unique and many of the errors could have been avoided with better communication and stronger staff development. The data indicated that few teachers received more than a basic overview of testing procedures consisting of distribution and collection of materials. Three recommendations were identified, based on this research, that require changes in policies and procedures. First, high schools are in need of designating a person responsible for all assessments. This person would become the site-testing expert and train staff members about the intricacies of instruments used. Second, the California Department of Education should require specific staff development with identified content centered around assessment. Third, assessment publishers, like Harcourt Brace, need to restructure the assessment materials to allow more flexibility for staff development using the "Directions for Administration." At this point, the "Directions for Administration" are secured documents. This prevents teachers from viewing them prior to testing, thus creating additional room for error.