1. Write a brief memo to Mary explaining decomposition of a time series.
2. Perform a multiplicative decomposition of total billable visits, save the residuals, and generate forecasts of visits for the next 12 months, using February FY2003-04 as the forecast origin.
3. Interpret the trend component for Mary. What did she learn from the seasonal indexes?
4. Compute the residual autocorrelations. Given the residual autocorrelations and the forecasts, should Mary be pleased with the decomposition analysis of total billable visits? Explain.
Mary Beasley and her goal of predicting the number of future total billable visits to Medical Oncology at Southwest Medical Center were introduced in Case 4-7. In that case, Mary learned there was a seasonal component in her data and considered using Winters' smoothing method to generate forecasts of future visits.