Forecasting Sales Volumes Flashcards
What is time series analysis?
A mathematical technique for analyzing past observations to forecast future impacts.
What are the four components of time series analysis?
- Long-term trend (T)
- Seasonal component (S)
- Cyclical variation (C)
- Random variation (R)
What does the long-term trend (T) represent?
The underlying direction and quantity of change over the long term.
What influences the long-term trend?
- Demographics
- Technological advancements
- Changes in lifestyles
Define a linear trend.
A straight line trend represented by the equation T = a + bt.
What is a logistic trend characterized by?
Slow growth followed by a spurt and then stagnation.
What is an exponential trend and how is it represented?
A growth pattern like compound interest, represented by T = (1 + r)^t.
Fill in the blank: Seasonal variations (S) are best described as _______.
[short-term trends and fluctuations due to seasonal circumstances]
What are cyclical variations (C)?
Variations that occur over a longer period, often associated with economic cycles.
True or False: Random variations (R) can be forecasted.
False
What is the easiest method to identify a trend?
Plotting observations on a graph and drawing a ‘line of best fit’.
What is an additive model in time series analysis?
Y = T + S + C + R, simplified to Y = T + S in exams.
How is the seasonal variation calculated in an additive model?
S = Y - T.
What does the multiplicative model represent?
Y = T x S x C x R, simplified to Y = T x S in exams.
How is the seasonal variation calculated in a multiplicative model?
S = Y / T.
Provide an example of seasonal variations.
- Higher sales of sun cream in summer
- Increased sales of scarves in winter
- Sales increases around holidays
What is the equation for exponential growth?
T = (1 + r)^t.
What is meant by adjusted values in time series analysis?
Forecasted figures that represent underlying trend values adjusted for seasonal variations.
What is the definition of adjusted values in forecasting?
Adjusted values represent the underlying trend values, adjusted for any seasonal variation.
In the equation y = 100 + 25x, what does ‘y’ represent?
Sales in units.
In the equation y = 100 + 25x, what does ‘x’ represent?
The quarter number.
How do you calculate the trend for a given quarter?
Set the x value equal to the quarter number and calculate the y value using the equation.
What is the formula for additive seasonal variation?
S = Y - T, where Y is actual sales and T is trend sales.
What is the additive variation for Quarter 1 if actual sales are 135 and the trend is 125?
10.
What is the formula for multiplicative seasonal variation?
S = Y / T.
What does a multiplicative seasonal variation of 1.08 indicate?
The number of units sold was 8% higher than the trend.
What are the forecast sales for Quarter 5 using the additive model?
235 units.
What are the forecast sales for Quarter 5 using the multiplicative model?
243 units.
True or False: The additive model assumes seasonal variations remain constant over time.
True.
True or False: The multiplicative model is generally considered superior to the additive model.
True.
What is the formula for the trend line in linear regression?
T = a + bt.
In the regression formula T = a + bt, what does ‘a’ represent?
The point at which the trend line intersects the vertical y-axis (at t = 0).
In the regression formula T = a + bt, what does ‘b’ represent?
The increase (or decrease) in one time period.
What does the symbol ∑ represent in regression analysis?
The ‘sum of’.
What is the least squares formula used for in linear regression?
To calculate the values of ‘a’ and ‘b’ for the trend line.
What is the time period represented by ‘t’ in the trend line formula?
The time period, starting from t = 1 for the first quarter.
What is the trend forecast for Q2 of 20X4 if t = 18?
72.38.
What would be the average of the y values (ȳ) if ∑y = 790.3 and n = 16?
49.39.
What would be the average of the x values (x̄) if ∑x = 136 and n = 16?
8.5.
What is the formula for calculating seasonal variation in the additive model?
Seasonal variation = Actual Value – Trend
What is the formula for calculating seasonal variation in the multiplicative model?
Seasonal variation = Actual Value / Trend
What are the average seasonal variations for each quarter?
- Q1: 0.83
- Q2: 1.00
- Q3: 0.96
- Q4: 1.21
What is the formula for forecasted values using the multiplicative model?
Actual (forecasted) values (Y) = Trend (T) x Seasonal variation (S)
What is the forecasted value for 20X4 Q1?
58.07
What is the forecasted value for 20X4 Q2?
72.38
What is the forecasted value for 20X4 Q3?
71.81
What is the forecasted value for 20X4 Q4?
93.44
What indicates a perfect correlation between two variables?
The line of best fit runs exactly through all of the points in the data set
What indicates a partial correlation between two variables?
A generally increasing or decreasing trend with some variation
What indicates no correlation between two variables?
No evidence of a relationship; points appear unrelated
What is the correlation coefficient represented by?
r
What does a correlation coefficient value close to 1 indicate?
A strong positive correlation
What does a correlation coefficient value close to 0 indicate?
No correlation between the variables
What does a correlation coefficient value close to -1 indicate?
A strong negative correlation
How is the coefficient of determination calculated?
Coefficient of determination = r²
What does a coefficient of determination of 0 indicate?
No relationship between changes in X and Y
What does a coefficient of determination of 1 indicate?
Change in Y is solely due to change in X
What is the rank correlation coefficient also known as?
Spearman’s rank correlation coefficient
What does the rank correlation coefficient measure?
The correlation between the rankings of two distributions
What is the formula for calculating the rank correlation coefficient?
Rs = 1 - (6 ∑d²) / (n(n² - 1))
What does a rank correlation coefficient close to 1 indicate?
A strong correlation between rankings
What can be concluded from a rank correlation coefficient of 0.636?
There is some correlation between productivity and happiness, but it is low