Chapter 3: Forecasting

The standard error of the estimate measures the scatter of the data points around the line

True

Forecast error is the difference between the actual and predicted values in a given period

True

Forecasts are the basis for virtually all decisions made in a business

True

Forecasts should be accurate

True

Which of the following are implications of accurate short-term forecasts?

- Greater credibility throughout the organization - Increased profits through operations - Increased confidence, which would allow the diversion of resources to longer-term planning

What are possible sources of forecast errors?

- Irregular variations - Random variation - Changes in the variables or relationships - Incorrect forecasting method or interpretation

Which of the following is/are disadvantages of the moving average forecast?

- It has acceptable error rates only for stable time series - You must find the value of n

What are advantages of the weighted moving average forecast?

- It is easy to compute - It gives more recent values higher weight - It is easy to understand

Qualitative forecasting approach

- Judgemental - Surverys

What are the differences of using the centered moving average method and the simple moving average method?

- The centered moving average works better when there is a trend in the data - The centered moving average typically requires more calculations

Which of the following statements about correlation is/are true?

- The closer to -1, the stronger the relationship is - The closer to +1, the stronger the relationship is - A correlation > 0 indicates a positive relationship

What are some disadvantages of the weighted moving average forecast?

- The value of n might be arbitrary - It requires considerable effort to determine the weights - It does not work well for time series with trends or seasonality

Which of the following statements are true about starting an exponential smoothing forecast?

- Use an average of several values to start off the initial forecast for the forecast series - Use a naive forecast for F2 to start off the forecast series

What statements are a correct characterization of salesforce opinions?

- While the customer may not be able to predict whether s/he will actually do what s/he wants, a salesperson can predict the follow-through - The salesforce is a good, unbiased source of information about customers' wants and needs

Forecast accuracy is important enough to take one's time to fine-tune the forecast as much as possible

False - A late forecast may not be useful anymore

A correlation of +1 is preferable to a correlation if -1

False - As long as the correlation is strong, the direction doesn't matter

Forecasts can be delivered either orally or in writing, as long as they are communicated clearly

False - In writing is very important, so they can be referred back to easily

Simple linear regression cannot be used if there is time-dependence in the data

False - It can be: simply use time as your predictor variable. This is identical to linear trend analysis

Accurate forecasts are easy to produce

False - It may be straightforward to do forecasts, but it is not necessarily the case. It can be especially difficult to provide accurate forecasts

Weighted moving average forecasts can be used for all types of time series

False - It should only be used for stable time series

The expected level (value) of the forecasted quantity is the most important aspect of the forecast

False - The degree of accuracy is also important

Long-term forecasts are used for ongoing operations

False - They are of no value for day-to-day operations

Simple linear regression can be done with relatively small amounts of data, e.g. n=5

False - You need at least 20 observations

Forecasts should be cost-effective

True

Forecasts should be in meaningful (understandable) units

True

Long-term forecasts pertain to items that will take a long while to implement

True

Naive forecasts are easy to understand

True

Naive forecasts can be used with stable time series, and with time series that have seasonality or trends

True

Naive forecasts use a single previous value to forecast a future value

True

Negative forecast errors occur when the forecast is too high

True

Positive forecast errors occur when the forecast is too low

True

Simple linear regression - if the correlation is small, the predictor variable is not much value and another variable may be better

True

Simple linear regression applies only when one predictor variable is used

True

Simple linear regression has all points carry equal weight

True

Simple linear regression is important to plot the data to see if a linear model is appropriate

True

Forecasts enable managers to plan for the future

True

A tracking signal compares the cumulative forecast error to the MAD in order to detect any ____ in errors over time

Bias

Quantitative forecasting approach

- Associative - Time- series

What is plotted on a control chart?

The errors

Which of the following are not components in time-series data?

- Competitor's actions - Outside forces - Market conditions

Deviations around the average value (i.e., the line) should be normally distributed. Which of the following supports the assumption of normality?

- A concentration of values close to the line - A small proportion of larger deviations

What are the similarities of using the centered moving average method and the simple moving average method?

- Both work well when there is no trend present in the data - Both compare a period average to find the seasonal relative or index

What is the correct formula for the linear trend equation?

F=a+bt

A correlation < 0 indicates a positive relationship

False

Salesforce opinions - The salesforce has direct contact with customers and therefore have a good sense of what customers want

False

Focus forecasting uses the forecasting technique that has the _____ accuracy for the given data set among a set of possible forecasting methods

Highest

The seasonal relative, also known as the seasonal _____, is the seasonal percentage applied in the ______ model

Index, multiplicative

The Delphi method is an _____ process which seeks to find a _______ forecast

Iterative, consensus

Bias is the persistent tendency of a forecast to be greater than or ____ than the ____ _______ of a time series

Less, actual values

Trend analysis develops an equation that will describe the trend present in data. The trend may be _____ or non-linear

Linear

Executive opinions are often used to develop _____ - ______ plans and ____ product development

Long-range, new

What is the correct interpretation for MAD?

MAD represents the average absolute forecast error

A value of 0.25 or less of r2 indicates a _____ predictor. A value between 0.25 and 0.8 indicates a _____ predictor

Poor, moderate

Predictor variables are variables that can be used to ______ _______ of the variable of interest

Predict values

Seasonal variation is:

Regularly repeating upward or downward movement that can be tied to recurring events

What type of correlation should there be between an indicator and the variable?

Strong

Correlation measures:

The strength and direction of the relationship between two variables

Time series date is a ____-______ sequence of observations taken at regular intervals

Time-ordered

Which of the following is the correct equation for the least squares regression line?

Yc=A+BX

The most common approach for forecasting cyclical date is explanatory, using ____

leads