Self-test Questions


1	Which of the following is not a basic component of classical time series analysis?

		A)	Economic variation
		B)	Seasonal variation
		C)	Cyclical variation
		D)	Random noise
		E)	Trend

2	Random error is the component of the classical time series models that the analyst is trying to eliminate.

		A)	TRUE
		B)	FALSE

3	In classical time series analysis, seasonal variation measures:

		A)	The effect of the weather on the data.
		B)	Regular, repeating patterns around the trend.
		C)	Long-term, irregular variations in the data.
		D)	The effect of growth across the years.
		E)	Short term random variation in the data.

4	What is usually the first step in carrying out a classical time series analysis with a time-based dataset?

		A)	Estimate the trend in the data.
		B)	Isolate the seasonal component.
		C)	Take a moving average of the data.
		D)	Choose whether to use a multiplicative or additive model.
		E)	Draw a scattergram of the data.

5	In a classical time series analysis the trend is found by:

		A)	Taking a moving average of the data points.
		B)	Correcting the moving average of the data points.
		C)	Estimating it from a scattergram.
		D)	Using least squares analysis.
		E)	Any of the above.

6	Why do we have to apply an adjustment to a 12 point moving average?

		A)	To minimise the effects of random error.
		B)	Because the exact midpoints of successive moving averages lie between existing data points.
		C)	To allow for the inherent instability of the process.
		D)	To ensure that seasonal variation is based on observations of all 12 periods.
		E)	To remove the trend component from the data.

7	What is usually the main consideration in deciding whether to use ‘ratio to trend’ or ’ratio to moving averages’ method to find seasonal indices?

		A)	The rate of trend.
		B)	Ease of calculation.
		C)	The degree of random variation.
		D)	Whichever minimises the mean square errors of the modelled values.
		E)	The relative importance of cyclical factors.

8	Classical time series analysis is a scientific method that always produces best-fit forecasts.

		A)	TRUE
		B)	FALSE

9	In general, what is the principal difference between moving average and exponential smoothing?

		A)	Moving averages are much easier to understand.
		B)	Moving averages require more calculations.
		C)	Exponential smoothing gives more emphasis to the most recent data points.
		D)	Exponential smoothing gives more accurate results than moving averages.
		E)	Past data points are given more weight by exponential smoothing.

10	In what situation might exponential smoothing be preferable to classical time series forecasting?

		A)	When you are short of time and need quick results.
		B)	When there is no obvious pattern in the data.
		C)	When you want to include all the data in the analysis.
		D)	When the trend is non-linear.
		E)	All of the above.

11	The Holt-Winters exponential smoothing model extends simple exponential smoothing to include trend and seasonal values.

		A)	TRUE
		B)	FALSE

12	If you were a manager in an ice cream company trying to forecast sales over the next year, using a large amount of historical data, which method is likely to give the best results?

		A)	Impossible to say.
		B)	Additive classical model.
		C)	Multiplicative classical model.
		D)	Simple exponential smoothing.
		E)	Holt-Winters exponential smoothing model.

13	The difference between successive points in a moving average is affected by the values of all the data points in the moving averages.

		A)	TRUE
		B)	FALSE

14	In the simple exponential smoothing model α is called the:

		A)	Smoothing constant
		B)	Stability parameter
		C)	Damping factor
		D)	Proportional exponential multiplier
		E)	First Greek letter

2003 A McGraw-Hill Online Learning Centre
Any use is subject to the Terms of Use and Privacy Policy.
McGraw-Hill Education Europe is one of the many fine businesses of The McGraw-Hill Companies.