Daily demand for marigold flowers at a large garden store is shown below:
Period Demand
1------------------- 85
2------------------- 92
3------------------- 71
4------------------ 97
5------------------- 93
6------------------ 82
7------------------- 89
1. Compute a 3-period moving average forecast and forecast error for each period. Then compute the cumulative forecast error.
2. Compute a 5-period moving average forecast and forecast error for each period. Then compute the cumulative forecast error.
3. Graph the demand data and the 3 and 5 period moving average forecasts on a graph.
4. Just observing the data line (not the forecasts), can you identify any trends, cycles, and/or seasonal patterns?
5. Using the cumulative forecast errors and any other assessment you wish, such as the pros and cons of shorter vs. longer averaging periods, the trends or cycles identified in #4, state which forecast you would recommend to use and why.
Your answer has to go beyond the generic information in the book that says shorter averaging periods respond more quickly to demand changes whereas longer periods smooth out peaks and valleys. You can certainly include that in your reasoning but then state why this is important to this type of business.