# Supply Chain Management Chapter 18

Flashcard maker : Lily Taylor
Continual review and updating in light of new data is a forecasting technique called second-guessing.
T
Cyclical influences on demand are often expressed graphically as a linear function that is either upward or downward sloping
F
Cyclical influences on demand may come from occurrences such as political elections, war, or economic conditions.
T
Trend lines are usually the last things considered when developing a forecast.
F
Time series forecasting models make predictions about the future based on analysis of past data.
T
In the weighted moving average forecasting model, the weights must add up to one times the number of data points.
F
In a forecasting model using simple exponential smoothing, the data pattern should remain stationary
T
In a forecasting model using simple moving average, the shorter the time span used for calculating the moving average, the closer the average follows volatile trends.
T
In the simple exponential smoothing forecasting model, you need at least 30 observations to set the smoothing constant alpha.
F
Experience and trial and error are the simplest ways to choose weights for the weighted moving average forecasting model.
T
Bayesian analysis is the simplest way to choose weights for the weighted moving average forecasting model.
F
The weighted moving average forecasting model uses a weighting scheme to modify the effects of individual data points. This is its major advantage over the simple moving average model.
T
A central premise of exponential smoothing is that more recent data are less indicative of the future than data from the distant past.
F
The equation for exponential smoothing states that the new forecast is equal to the old forecast plus the error of the old forecast.
F
Exponential smoothing is always the best and most accurate of all forecasting models.
F
In exponential smoothing, it is desirable to use a higher smoothing constant when forecasting demand for a product experiencing high growth.
T
The value of the smoothing constant alpha in an exponential smoothing model is between 0 and 1.
T
Simple exponential smoothing lags changes in demand.
T
Exponential smoothing forecasts always lag behind the actual occurrence but can be corrected somewhat with a trend adjustment
T
Because the factors governing demand for products are very complex, all forecasts of demand contain error.
T
Random errors can be defined as those that cannot be explained by the forecast model being used.
T
There are no differences in strategic and tactical forecasting. A forecast is a mathematical projection, and its ultimate purpose should make no difference to the analyst.
F
Random errors in forecasting occur when an undetected secular trend is not included in a forecasting model
F
When forecast errors occur in a normally distributed pattern, the ratio of the mean absolute deviation to the standard deviation is 2 to 1, or 2 x MAD = 1 standard deviation
F
When the errors that occur in the forecast are normally distributed (the usual case,) the mean absolute deviation (MAD) relates to the standard deviation as 1 standard deviation = 1.25 MAD.
MAD statistics can be used to generate tracking signals
T
RSFE in forecasting stands for “reliable safety function error.”
F
In forecasting, RSFE stands for “running sum of forecast errors.”
T
A tracking signal (TS) can be calculated using the arithmetic sum of forecast deviations divided by the MAD.
T
A restriction in using linear regression is that it assumes that past data and future projections fall on or near a straight line.
T
Regression is a functional relationship between two or more correlated variables, where one variable is used to predict another.
T
Linear regression is not useful for aggregate planning.
F
The standard error of the estimate of a linear regression is not useful for judging the fit between the data and the regression line when doing forecasts.
F
Multiple regression analysis uses several regression models to generate a forecast.
F
For every forecasting problem, there is one best forecasting technique.
F
A good forecaster is one who develops special skills and experience at one forecasting technique and is capable of applying it to widely diverse situations.
F
In causal relationship forecasting, leading indicators are used to forecast occurrences.
F
Qualitative forecasting techniques generally take advantage of the knowledge of experts and therefore do not require much judgment.
F
Market research is a quantitative method of forecasting
F
Decomposition of a time series means identifying and separating the time series data into its components.
T
A time series is defined in the text as chronologically ordered data that may contain one or more components of demand variation: trend, seasonal, cyclical, autocorrelation, and random.
T
It is difficult to identify the trend in time series data.
F
In decomposition of time series data, it is relatively easy to identify cycles and autocorrelation components.
F
We usually associate the word “seasonal” with recurrent periods of repetitive activity that happen on other than an annual cycle.
F
In time series data depicting demand, which of the following is not considered a component of demand variation?

A. Trend

B. Seasonal

C. Cyclical

D. Variance

E. Autocorrelation

D
Which of the following is not one of the basic types of forecasting?

A. Qualitative

B. Time series analysis

C. Causal relationships

D. Simulation

E. Force field analysis

E
In most cases, demand for products or services can be broken down into several components. Which of the following is not considered a component of demand?

A. Average demand for a period

B. A trend

C. Seasonal elements

D. Past data

E. Autocorrelation

D
In most cases, demand for products or services can be broken into several components. Which of the following is considered a component of demand?

A. Cyclical elements

B. Future demand

C. Past demand

D. Inconsistent demand

E. Level demand

A
In most cases, demand for products or services can be broken into several components. Which of the following is considered a component of demand?

A. Forecast error

B. Autocorrelation

C. Previous demand

D. Consistent demand

E. Repeat demand

B
Which of the following forecasting methodologies is considered a qualitative forecasting technique?

A. Simple moving average

B. Market research

C. Linear regression

D. Exponential smoothing

E. Multiple regression

B
Which of the following forecasting methodologies is considered a time series forecasting technique?

A. Simple moving average

B. Market research

D. Historical analogy

E. Simulation

A
Which of the following forecasting methodologies is considered a time series forecasting technique?

A. Delphi method

B. Exponential averaging

C. Simple movement smoothing

D. Weighted moving average

E. Simulation

D
Which of the following forecasting methodologies is considered a causal forecasting technique?

A. Exponential smoothing

B. Weighted moving average

C. Linear regression

D. Historical analogy

E. Market research

C
Which of the following forecasting methods uses executive judgment as its primary component for forecasting?

A. Historical analogy

B. Time series analysis

C. Panel consensus

D. Market research

E. Linear regression

C
Which of the following forecasting methods is very dependent on selection of the right individuals who will judgmentally be used to actually generate the forecast?

A. Time series analysis

B. Simple moving average

C. Weighted moving average

D. Delphi method

E. Panel consensus

B
In business forecasting, what is usually considered a short-term time period?

A. Four weeks or less

B. More than three months

C. Six months or more

D. Less than three months

E. One year

D
In business forecasting, what is usually considered a medium-term time period?

A. Six weeks to one year

B. Three months to two years

C. One to five years

D. One to six months

E. Six months to six years

B
In business forecasting, what is usually considered a long-term time period?

A. Three months or longer

B. Six months or longer

C. One year or longer

D. Two years or longer

E. Ten years or longer

D
In general, which forecasting time frame compensates most effectively for random variation and short-term changes?

A. Short-term forecasts

B. Quick-time forecasts

C. Long-range forecasts

D. Medium-term forecasts

E. Rapid-change forecasts

A
In general, which forecasting time frame best identifies seasonal effects?

A. Short-term forecasts

B. Quick-time forecasts

C. Long-range forecasts

D. Medium-term forecasts

E. Rapid-change forecasts

D
In general, which forecasting time frame is best to detect general trends?

A. Short-term forecasts

B. Quick-time forecasts

C. Long-range forecasts

D. Medium-term forecasts

E. Rapid-change forecasts

C
Which of the following forecasting methods can be used for short-term forecasting?

A. Simple exponential smoothing

B. Delphi technique

C. Market research

D. Hoskins-Hamilton smoothing

E. Serial regression

A
Which of the following considerations is not a factor in deciding which forecasting model a firm should choose?

A. Time horizon to forecast

B. Product

C. Accuracy required

D. Data availability

E. Analyst availability

B
A company wants to forecast demand using the simple moving average. If the company uses four prior yearly sales values (i.e., year 2010 = 100, year 2011 = 120, year 2012 = 140, and year 2013 = 210), which of the following is the simple moving average forecast for year 2014?

A. 100.5

B. 140.0

C. 142.5

D. 145.5

E. 155.0

C Forecast for 2014 = (100 + 120 + 140 + 210)/4 = 570/4 = 142.5.
A company wants to forecast demand using the simple moving average. If the company uses three prior yearly sales values (i.e., year 2011 = 130, year 2012 = 110, and year 2013 = 160), which of the following is the simple moving average forecast for year 2014?

A. 100.5

B. 122.5

C. 133.3

D. 135.6

E. 139.3

C Forecast for 2014 = (130 + 110 + 160)/3 = 400/4 = 133.3.
A company wants to forecast demand using the weighted moving average. If the company uses two prior yearly sales values (i.e., year 2012 = 110 and year 2013 = 130), and we want to weight year 2012 at 10 percent and year 2013 at 90 percent, which of the following is the weighted moving average forecast for year 2014?

A. 120

B. 128

C. 133

D. 138

E. 142

B Forecast for 2014 = (160 x 0.3) + (140 x 0.3) + (170 x 0.4) = 158.
A company wants to forecast demand using the weighted moving average. If the company uses three prior yearly sales values (i.e., year 2011 = 160, year 2012 = 140, and year 2013 = 170), and we want to weight year 2011 at 30 percent, year 2012 at 30 percent, and year 2013 at 40 percent, which of the following is the weighted moving average forecast for year 2014?

A. 170

B. 168

C. 158

D. 152

E. 146

C Forecast for 2014 = (160 x 0.3) + (140 x 0.3) + (170 x 0.4) = 158.
The exponential smoothing method requires which of the following data to forecast the future?
A
Given a prior forecast demand value of 230, a related actual demand value of 250, and a smoothing constant alpha of 0.1, what is the exponential smoothing forecast value for the following period?

A. 230

B. 232

C. 238

D. 248

E. 250

230 + 0.1 x (250 – 230) = 232
If a firm produced a standard item with relatively stable demand, the smoothing constant alpha (reaction rate to differences) used in an exponential smoothing forecasting model would tend to be in which of the following ranges?

A. 5 to 10 percent

B. 20 to 50 percent

C. 20 to 80 percent

D. 60 to 120 percent

E. 90 to 100 percent

A
If a firm produced a product that was experiencing growth in demand, the smoothing constant alpha (reaction rate to differences) used in an exponential smoothing forecasting model would tend to be which of the following?

A. Close to zero

B. A very low percentage, less than 10 percent

C. The more rapid the growth, the higher the percentage

D. The more rapid the growth, the lower the percentage

E. 50 percent or more

C
Given a prior forecast demand value of 1,100, a related actual demand value of 1,000, and a smoothing constant alpha of 0.3, what is the exponential smoothing forecast value?

A. 1,000

B. 1,030

C. 1,070

D. 1,130

E. 970

B 1,100 + 0.3 x (1,100 – 1,000) = 1,030.
A company wants to generate a forecast for unit demand for year 2014 using exponential smoothing. The actual demand in year 2013 was 120. The forecast demand in year 2013 was 110. Using these data and a smoothing constant alpha of 0.1, which of the following is the resulting year 2014 forecast value?

A. 100

B. 110

C. 111

D. 114

E. 120

C
As a consultant, you have been asked to generate a unit demand forecast for a product for year 2014 using exponential smoothing. The actual demand in year 2013 was 750. The forecast demand in year 2013 was 960. Using these data and a smoothing constant alpha of 0.3, which of the following is the resulting year 2014 forecast value?

A. 766

B. 813

C. 897

D. 1,023

E. 1,120

C
Which of the following is a possible source of bias error in forecasting?

A. Failing to include the right variables

B. Using the wrong forecasting method

C. Employing less sophisticated analysts than necessary

D. Using incorrect data

E. Using standard deviation rather than MAD

A
Which of the following is used to describe the degree of error?

A. Weighted moving average

B. Regression

C. Moving average

D. Forecast as a percent of actual

E. Mean absolute deviation

E
A company has actual unit demand for three consecutive years of 124, 126, and 135. The respective forecasts for the same three years are 120, 120, and 130. Which of the following is the resulting MAD value that can be computed from these data?

A. 1

B. 3

C. 5

D. 15

E. 123

C MAD = ABS[(124 – 120) + (126 – 120) + (135 – 130)]/3 = 15/3 = 5.
A company has actual unit demand for four consecutive years of 100, 105, 135, and 150. The respective forecasts were 120 for all four years. Which of the following is the resulting MAD value that can be computed from these data?

A. 2.5

B. 10

C. 20

D. 22.5

E. 30

C MAD = ABS[(100 – 120) + (105 – 120) + (135 – 120) + (150 – 120)]/4 = 80/4 = 20.
If you were selecting from a variety of forecasting models based on MAD, which of the following MAD values from the same data would reflect the most accurate model?

A. 0.2

B. 0.8

C. 1.0

D. 10.0

E. 100.0

A
A company has calculated its running sum of forecast errors to be 500, and its mean absolute deviation is exactly 35. Which of the following is the company’s tracking signal?

A. Cannot be calculated based on this information

C. More than 35

D. Exactly 35

B The tracking signal is RSFE/MAD = 500/35 = 14.29.
A company has a MAD of 10. It wants to have a 99.7 percent control limit on its forecasting system. Its most recent tracking signal value is 3.1. What can the company conclude from this information?

A. The forecasting model is operating acceptably.

B. The forecasting model is out of control and needs to be corrected.

C. The MAD value is incorrect.

D. The upper control value is less than 20.

E. It is using an inappropriate forecasting methodology.

A Tracking signal = RSFE/MAD; hence, 3.1 = RSFE/10 or RSFE = 3.1 x 10 = 31. MAD = 10, SD = 1.25 x MAD = 12.5. Because 99.7 percent corresponds to 3 standard deviations from the mean, RSFE would have to be higher than 3 x 12.5 or 37.5 for the forecasting model to be out of control.
You are hired as a consultant to advise a small firm on forecasting methodology. Based on your research, you find the company has a MAD of 3. It wants to have a 99.7 percent control limits on its forecasting system. Its most recent tracking signal value is 15. What should be your report to the company?
A. The forecasting model is operating acceptably.

B. The forecasting model is out of control and needs to be corrected.

C. The MAD value is incorrect.

D. The upper control value is less than 20.

E. The company is using an inappropriate forecasting methodology.

B Tracking signal = RSFE/MAD; hence, 15 = RSFE/3 or RSFE = 15 x 3 = 45. MAD = 3, SD = 1.25 x MAD = 3.75. Because 99.7 percent corresponds to 3 standard deviations from the mean, 3 x 3.75 = 11.25. Because RSFE is 45, the forecasting model is out of control.
Which of the following is the percentage of observations you would expect to see lying within a plus or minus 3 MAD range?

A. 57.05 percent

B. 88.95 percent

C. 98.36 percent

D. 99.85 percent

E. 100 percent

C 3 MAD x 0.8 = 2.4 standard deviations. From Appendix E, 2.4 standard deviations includes 0.4918 of the area x 2 = 0.9836, or 98.36 percent.
Which of the following is the percentage of observations you would expect to see lying within a plus or minus 2 MAD range?

A. 57.04 percent

B. 89.04 percent

C. 98.33 percent

D. 99.86 percent

E. 100.00 percent

B 2 MAD x 0.8 = 1.6 standard deviations. From Appendix E, 1.6 standard deviations = 0.4452 of the area x 2 = 0.8904, or 89.04 percent
If the intercept value of a linear regression model is 40, the slope value is 40, and the value of X is 40, which of the following is the resulting forecast value using this model?

A. 120

B. 1,600

C. 1,640

D. 2,200

E. 64,000

C The linear regression line is of the form Y = a + bX, where Y is the value of the dependent variable that we are solving for, a is the Y intercept, b is the slope, and X is the independent variable. Hence, Y = 40 + 40 x 40 = 1,640.
A company hires you to develop a linear regression forecasting model. Based on the company’s historical sales information, you determine the intercept value of the model to be 1,200. You also find the slope value is minus 50. If, after developing the model, you are given a value of X = 10, which of the following is the resulting forecast value using this model?

A. -1,800

B. 700

C. 1,230

D. 1,150

E. 12,000

B The linear regression line is of the form Y = a + bX, where Y is the value of the dependent variable that we are solving for, a is the Y intercept, b is the slope, and X is the independent variable. Hence, Y = 1,200 + (-50) x 10 = 700.
Heavy sales of umbrellas during a rain storm is an example of which of the following?

A. A trend

B. A causal relationship

C. A statistical correlation

D. A coincidence

B
You are using an exponential smoothing model for forecasting. The running sum of the forecast error statistics (RSFE) are calculated each time a forecast is generated. You find the last RSFE to be 34. Originally, the forecasting model used was selected because of its relatively low MAD of 0.4. To determine when it is time to re-evaluate the usefulness of the exponential smoothing model, you compute tracking signals. Which of the following is the resulting tracking signal?

A. 85

B. 60

C. 13.6

D. 12.9

E. 8

A 340/.4