1. Develop a multiple regression equation using unemployment rate and average monthly temperature to predict emergency road service calls.
2. Create a new temperature variable and relate it to emergency road service. Remember that temperature is a relative scale and that the selection of the zero point is arbitrary. If vehicles are designed to operate best at 65 degrees Fahrenheit, then every degree above or below 65 degrees should make vehicles operate less reliably. To accomplish a transformation of the temperature data that simulates this effect, begin by subtracting 65 from the average monthly temperature values. This repositions "zero" to 65 degrees Fahrenheit. Should absolute values of this new temperature variable be used?
3. Create a new unemployment rate variable and relate it to emergency road service. Give unemployment a lagged effect on emergency road service by using the unemployment rate for the month (1) three months prior to the current month and (2) 11 months prior to the current month as the data for the unemployment independent variable. Which model is better for prediction? Are the signs on the coefficients for the independent variables what you would expect them to be? Are the coefficients of the independent variables significantly different from zero?
4. Develop a multiple regression equation using the transformed average temperature variable created in step 2 and the lagged unemployment variable created in step 3. Is this a good model? Have any of the underlying assumptions been violated?
An overview of AAA Washington was provided in Case 5-5 when students were asked to prepare a time series decomposition of the emergency road service calls received by the club over five years. The analysis performed in Case 5-5 showed that the pattern in emergency road service call volume was probably somewhat cyclical in nature. In Case 6-6, four variables were investigated: unemployment rate, average daily temperature, amount of rainfall, and number of members in the club.