# Assignment 3 2

**Topics:**Regression analysis, Forecasting, Linear regression, Time series analysis /

**Pages:**5 (773 words) /

**Published:**Apr 4th, 2015

1. The Accuweather Corporation manufactures barometers and thermometers for weather forecasters. In an attempt to forecast its future needs for mercury, Accuweather's chief economist estimated average monthly mercury needs as:

N = 500 + 10X

where N = monthly mercury needs (units) and X = time period in months (January 2008= 0). The following monthly seasonal adjustment factors have been estimated using data from the past five years:

Month

Adjustment Factor

January

15%

April

10%

July

20%

September

5%

December

10%

(a)

Forecast Accuweather's mercury needs for January, April, July, September, and December of 2010. (1)

(b)

The following actual and forecast values of mercury needs in the month of November have been recorded:

Year

Actual

Forecast

2008

456

480

2009

324

360

2007

240

240

What seasonal adjustment factor should the firm use for November? (1)

2. The variation in an economic time-series which is caused by major expansions or contractions usually (1) of greater than a year in duration is known as:

a.

secular trend

b.

cyclical variation

c.

seasonal effect

d.

unpredictable random factor

e.

none of the above

3. Examine the plot of data. (1) Sales

Time

It is likely that the best forecasting method for this plot would be:

a. a two-period moving average

b. a secular trend upward

c. a seasonal pattern that can be modeled using dummy variables or seasonal adjustments

d. a semi-log regression model

e. a cubic functional form

4. Emma uses a linear model to forecast quarterly same-store sales at the local Garden Center. The results of her multiple regression is: (1 point)

Sales = 2,800 + 200•T - 350•D

where T goes from 1 to 16 for each quarter of the year from the first quarter of 2006 (‘06I) through the fourth quarter of 2009 (‘09 IV). D is a dummy variable which is 1 if sales are in the cold and dreary first quarter, and zero