No substantive changes have been made to the original time study proposal (Spring 2004). We have only updated references to numbers of districts and staff participating. We have attached the most recent forms, definitions, sampling, monitoring and training procedures, contracts, and detailed data. A modified Random Moment Sampling (RMS) design is planned, where both staff and time are randomly sampled, and the frequency of certain activities is measured. Staff samples provide time estimates based on 5% error and 95% confidence. This RMS design makes conservative assumptions and aims to satisfy desired levels of statistical validity. Since Medicaid reimbursable activities represent a very small proportion of most school staff time – about 1-4% of the total work day, purely random time moments would capture inapplicable activities 96-99% of the time. Surveying staff at any random moment concerning their activities would be quite distracting for most respondents (while teachers are teaching, counselors are counseling, etc.) and a very inefficient way of collecting such information.
