Question

Do you have a favorite statistical method?

Is there a statistical method that you think you will/already use in your workplace? v

Answer 1

My favourite statistical method is Linear Regression.

Through regression I can try to predict future sales/revenue based on past data. I also implore time series forecasting to compensate for the seasonality present in the data.

Example

I have Revenue data for 5 Weeks (Only 6 days in week are considered as the restaurant is closed on every Sunday)

We will be calculating the seasonality of each day (ie Mon – Sat) and use it to predict Revenue for Week 6 (Mon-Sat)

Steps:

1. First we calculate 4 Period(Days) moving Average.
2. Then we calculate Centred Average of these 4 Period moving Average.
3. Then % Average ie Revenue/Centred Average
4. Then we table the % Average for each day (1-12) for the five weeks and calculate mean for each day. (As Shown in Table 2)
5. 600%/The total of these 6 Period (Days) mean will be the x Adjustment Factor
6. Seasonality Index for each day will be X Adj Factor * Day’s Mean ie Seasonality Index for Monday = 66.8%*8.98 = 6.00
7. Then we calculate ∑X, ∑Y, ∑XY, ∑X² and using these values we calculate the regression equation.
8. Regression Equation will be: Y = 1442.83 + 14.84*X
9. Using this equation and Seasonality Index, we calculate value for Predicted Revenue for next 6 Days. Eg: Monday (Week 6) Revenue = (1442.83 + 14.84*31)*6

Table 1

Table 2

Using the above method we predict sales for Week 6