Deck 11: Forecasting and Demand Planning

Question 1

Top managers use detailed forecasts of unit sales for individual products (e.g., brands and sizes) for decisions involving financial planning and for sizing and locating new facilities.

Answer 1

False

Question 2

Exponential smoothing models "never forget" past data as long as the smoothing constant is strictly between 0 and 1. In contrast, moving average methods "completely forget" all data older than k periods in the past.

Answer 2

True

Question 3

Because of random variations, forecasts are never 100% accurate.

Answer 3

True

Question 4

Better operational decisions can be made by integrating forecasting with value chain and capacity management systems.

Question 5

Aggregate forecasts are generally much easier to develop whereas detailed forecasts require more time and resources.

Question 6

Long-range forecasts expressed in sales dollars are more meaningful to top managers than to managers at the operations level.

Question 7

In a regression model, both the dependent and independent variables must be numerical.

Question 8

All time series contain random variation but may not contain trend or seasonal components.

Question 9

An R² of 0.70 mean 30% of the variability in the dependent variable was explained by the independent variable.

Question 10

Another name for planning horizon is time bucket.

Question 11

Seasonal patterns can occur over the weeks during a month, over days during a week, or hours during a day.

Question 12

A single moving average is most appropriate for data with identifiable trends.

Question 13

A major difference between MSE and MAD is that MAD is influenced much more by large forecast errors than by small errors.

Question 14

A long-range forecast typically covers a planning horizon of 3 to 12 months.

Question 15

Statistical forecasting is based upon the assumption that the future will be an extrapolation of the past.

Question 16

In forecasting, irregular variation that is explainable can normally be discarded.

Question 17

MAD, MSE, and MAPE forecast error metrics generally give similar numerical results so it doesn't matter which one is used.

Question 18

As the value of k is increased in a moving average forecasting model, the forecast reacts more slowly to recent changes in the time series.

Question 19

In an exponential smoothing model, larger values of alpha place less emphasis on recent data and more on older data.

Question 20

Trends are characterized by repeatable periods of ups and downs over short periods of time.

Question 21

Regular patterns in a data series that take place over long periods of time are called ____.

A)Trends
B)Seasonal patterns
C)Cyclical patterns
D)Irregular variation

Question 22

The Delphi method is a forecasting approach that is based on expert opinion.

Question 23

____ forecasts are needed for planning production schedules and to assign workers to jobs.

A)Long-range
B)Intermediate-range
C)Short-range
D)Demand planning

Question 24

In practice, managers rely almost exclusively on statistical forecasts.

Question 25

A(n) ____ is a one-time variation that is explainable.

A)Cyclical pattern
B)Random Variation
C)Irregular variation
D)Seasonal pattern

Question 26

The smoothing constant, $\alpha$ , used in the basic exponential smoothing model, can range in value from -1 to +1.

Question 27

An exponential smoothing model can be found easily by applying the Excel Add Trendline option to a time series.

Question 28

The forecasting error measurement that is different in that the measurement scale factor is eliminated is

A)MSE
B)MAD
C)RMSE
D)MAPE

Question 29

Regression models are often used in forecasting to incorporate causal variables that may influence a time series.

Question 30

Irregular variation and random variation both refer to unexplainable deviation of a time series from a predictable pattern.

Question 31

Repeatable periods of ups and downs over short periods of time are called ____.

A)Trends
B)Seasonal patterns
C)Cyclical patterns
D)Irregular variation

Question 32

Single exponential smoothing with a smoothing constant value of .25 puts the same weight on the most recent actual demand as a 4-period moving average.

Question 33

Which of the following is not a statistical method?

A)Delphi
B)Exponential smoothing
C)Moving average
D)Linear regression

Question 34

____ forecasts are needed to plan work-force levels, allocate budgets among divisions and schedule jobs and resources.

A)Long-range
B)Intermediate-range
C)Short-range
D)Demand planning

Question 35

Judgmental forecasting should only be used if no historical data are available.

Question 36

A major difference between forecast accuracy measures MAD and MSE is that MAD is influenced much more by large forecast errors than by small errors.

Question 37

A moving average model works best when ____ in the time series.

A)Only irregular variation is present
B)Only a trend is present
C)There is no trend, seasonal, or cyclical pattern
D)Trend, seasonal, and cyclical patterns all exist

Question 38

Single exponential smoothing is a forecasting technique that uses a weighted average of past time-series values to forecast the value of the time series in the next period.

Question 39

____ forecasts are needed to plan for facility expansion.

A)Long-range
B)Intermediate-range
C)Short-range
D)Demand planning

Question 40

Which of the following is not one of the five characteristics of a time series?

A)Time bucket
B)Trend
C)Cyclical
D)Random variation

Question 41

Which of the following is not a valid approach to gathering data for judgmental forecasting?

A)Questionnaire
B)Telephone contact
C)Personal interview
D)Company records

Question 42

Define forecast error and describe ways that it is measured.

Question 43

Define regression analysis and explain how it is an approach to forecasting.

Question 44

Which is not true regarding simple exponential smoothing?

A)Small values of alpha ( $\alpha$ ) place more emphasis on past data
B)Larger values of alpha ( $\alpha$ ) have the advantage of quickly adjusting the forecast
C)If alpha ( $\alpha$ ) equals zero, the forecast will never change
D)If alpha ( $\alpha$ ) equals one, the forecast will never change

Question 45

An R² of 0.80 means

A)80% of the variability in the independent variable is explained by the dependent variable
B)80% of the variability in the dependent variable is explained by the independent variable
C)80% of the variability in the dependent variable is not explained by the independent variable
D)Multiple regression was used

Question 46

If single exponential smoothing is used and the time series has a negative trend, the forecast will

A)Lag
B)Overshoot
C)Be on target
D)Have a MAD equal to zero

Question 47

Explain how forecasting is used at different levels of an organization.

Question 48

Discuss the three planning horizons used in forecasting and the types of decisions made in each.

Question 49

Regression analysis

A)Is limited to one dependent and one independent variable
B)Is best with linear relationships
C)Maximizes the sum of the squared deviations between the actual time series value and the estimated values of the dependent variable
D)Can be used with time as the independent variable

Question 50

All of the following are important in choosing a forecasting method except

A)Smoothing constant ( $\alpha$ )
B)Time span for which forecast is made
C)Data requirements
D)Quantitative skills needed

Question 51

Exponential smoothing...

A)Works best for long-term forecasting
B)Yields a mathematically optimal solution
C)Assigns weights to past data that decay exponentially as the data gets older
D)Cannot be adapted to handle trend

Question 52

A tracking signal provides a method for quantifying forecast

A)bias
B)error
C)accuracy
D)outliers

Question 53

Explain the difference between statistical forecasting and judgment forecasting.

Question 54

Explain the difference between a moving average and single exponential smoothing forecasting model.

Question 55

What is a time series, and what types of characteristics typically make up time series?

Question 56

All of the following are important concepts in forecasting except

A)Determining the planning horizon length.
B)Determining the time bucket size (i.e., year, quarter, month, week, day, etc.).
C)Using a smoothing constant of 0.1 in Delphi methods of forecasting.
D)Identifying cyclical patterns.

Question 57

If actual demand for a product is highly influenced by only random variation, the quantitative technique to use for forecasting demand is

A)Regression
B)Moving average
C)Mean Absolute Percentage Error (MAPE)
D)Delphi

Question 58

A group of international experts published a set of principles of forecasting that includes all of the following except

A)Use quantitative rather than qualitative methods.
B)Combine forecasts from approaches that are similar.
C)Ask experts to justify their forecasts in writing.
D)Use multiple measures of forecast accuracy.

Question 59

For single exponential smoothing,

A)Large values of alpha ( $\alpha$ ) place more emphasis on recent data
B)Small values of alpha ( $\alpha$ ) place more emphasis on recent data
C)Very volatile time series with substantial random variability should use a large value for alpha ( $\alpha$ )
D)Very stable time series with little random variability should use small values for alpha ( $\alpha$ )

Question 60

Which of the following does not fit with the Delphi method?

A)Group of experts
B)Brought together as a group
C)Process iterates until a consensus is reached
D)Qualitative as well as numerical outputs

Question 61

A 7-month simple moving average would approximately equate with what alpha ( $\alpha$ ) factor for simple exponential smoothing?

Question 62

Sales of a new CD at a store for the last 4 weeks are shown below.
a.Find a 3 period moving average forecast for the next week.
b.Find a 4 period moving average forecast for the next week.
c.Actual sales for week 5 were 105 units. What would be the 3- and 4-period moving average forecasts for week 6?

Question 63

The Espresso Cart has had the following pattern of espresso sales over the last two weeks: What is the forecast for Friday's sales using a three-day moving average?

Question 64

A Taiwan electronics company exports personal computers (PCs) to the U.S. Their PC sales (in thousands) over the past five years are given below.
a.What is the regression equation if the company wants to predict sales?
b.What is the forecast for sales in year 6?

Question 65

Explain judgmental forecasting, including grass roots forecasting and the Delphi Method.

Question 66

Describe some of the issues that managers must consider in applying forecasting methods in practice.

Question 67

Ed Rogers owns an appliance store. Sales data on a particular model of a DVD player for the past six months are:
Forecast sales for July using an exponential smoothing model with a smoothing constant of 0.40. Assume that the forecast for May was 36.25.

Question 68

Valentine's Day is the best day of the year for selling roses at River Road Florist. Dozens of roses sold on Valentine's Day over six years are as follows:
a.What is the forecast for Valentine's Day in 2011 using a 3-period moving average?
b.What is the forecast for Valentine's Day in 2011 using a 5-period moving average?
c.What is the MAD for years 2008-2010 using a 3-period moving average?
d.What is the tracking signal for years 2008-2010 using a 3-period moving average?

Question 69

Consider the sales for six consecutive weeks for Sam's Strawberries. The sales are in "flats" sold.
a.Using a 3-period moving average, forecast the sales for weeks four through six.
b.Use exponential smoothing with $\alpha$ = .3 to forecast sales for weeks four through six.
c.Use linear regression (time series) to develop a prediction equation that will forecast sales. Then use that prediction equation to get forecasts for weeks four through six.
d.Use MAD to pick the best forecasting method of A through C.

Question 70

Ed Rogers owns an appliance store. Sales data on a particular model of a DVD player for the past six months are shown below along with the results of two different forecasting models that were developed. Which is the better forecasting model, based on the MAD criterion?

Question 71

Using the data shown below, compute the mean square error, mean absolute deviation, and mean absolute percentage error for the forecasts shown.

Question 72

The following data represents the home mortgage loan interest rates at a local bank over an eight-month period:
a.What is the forecast for month 8 using a moving average model with an AP = 4?
b.What is the forecast for month 9 using a moving average model with an AP = 6?

Question 73

Community General Hospital finds itself treating many bicycle accident victims. Data from the last seven 24-hour periods is shown below:
a.What is the forecast for day 4 using a moving average model with AP = 3?
b.With an alpha value of .5 and a starting forecast in day 4 equal to the actual data, what is the exponentially smoothed forecast for day 8?
c.What is the MAD for days 6 to 8 for an exponentially smoothed forecasting model with an alpha value of .5 and a starting forecast in day 4 equal to the actual data?
d.What is the tracking signal for days 6 to 8 for an exponentially smoothed forecasting model with an alpha value of .5 and a starting forecast in day 4 equal to the actual data?

Question 74

State Division of Motor Vehicles (DMV) statistics show the rate of new driver's license applications to be as shown below:
a.Using a 3-week moving average, what is the forecast for the first week in April?
b.Using a 5-week moving average, what is the forecast for the first week in April?

Question 75

The manager of a gas station along an interstate highway has observed that gasoline sales generally increase each week over the summer months as more families travel by car on vacations. He also believes that sales are sensitive to fluctuations in the price of gasoline. He developed the following regression model:
Sales = 59407 + 509 (Week) -16463 (Price/gallon)
a.Interpret the coefficients of the independent variables in this model.
b.What is the sales forecast for the 11^th week of the summer if the price per gallon is estimated to be $3.00?

Question 76

Based on the information shown below, develop forecasts for June using both a 2-period moving average model and an exponential smoothing model with $\alpha$ = 0.10. For the exponential smoothing model, assume the forecast for February was 800.